Opinion
5:21-cv-00418-RDP
12-10-2024
CENT HOLDING COMPANY, LLC, et al., Plaintiffs, v. WOLFRAM RESEARCH INC., Defendant.
MEMORANDUM OPINION
R. DAVID PROCTOR, CHIEF U.S. DISTRICT JUDGE
This matter is before the court on Defendant Wolfram Research, Inc.'s Motion for Summary Judgment (Doc. # 163), and Motion to Bar the Expert Testimony of W. Anthony Mason (Doc. # 161). The parties have fully briefed the Motion for Summary Judgment (Docs. # 164, 166, 178, 181), as well as Defendant's Motion to Bar Expert Testimony (Docs. # 173, 174). For the reasons explained below, Defendant's Motion for Summary Judgment is due to be granted, and its Motion to Bar Expert Testimony is due to be denied. Below, the court outlines the relevant background information and procedural history of this case. The court then discusses the undisputed facts, the standard of review, and how the facts apply to Plaintiffs' legal claims.
I. Background and Procedural History
This is a trade secrets case. Although the facts are riddled with technical jargon, the allegation is relatively straightforward: Plaintiffs allege that Defendant wrongfully appropriated and publicized their computer code. Plaintiffs say this code is especially valuable because it could aid Plaintiffs (and those to whom they sell the product) in beating the stock market. They contend that by downloading and processing stock market data faster and at a higher volume than any other code, their code holds the potential for getting ahead of market developments and acquiring significant wealth. Defendant counters that it did not steal or publicize any of Plaintiffs' code or trade secrets. Rather, Defendant argues, any similarity between Defendant's code at issue here and Plaintiffs' code is irrelevant (1) because Defendant was already using the code at issue before meeting Plaintiffs, and, even more significantly, (2) because Plaintiffs have not pointed to any specific trade secret that Defendant stole and publicized.
Plaintiff Cent Capital, LLC is a start-up hedge fund formed in 2020 with the goal of using computer science to analyze securities data to make better investment decisions for its clients. (Doc. # 125 ¶ 19). Plaintiff Cent Holding Company, LLC owns 100% of Cent Capital, LLC. (Id.). The members of both companies reside in and are citizens of Alabama. (Id. ¶¶ 12-13). Defendant Wolfram Research Inc. is a company based in Illinois that develops and sells computer languages and programs for use in computer coding. (Id. ¶¶ 14, 23). Plaintiffs allege that Defendant stole Plaintiffs' trade secrets (specifically, certain portions, combinations of, and ideas for computer code), disclosed these secrets in at least one public webinar, and improperly used these secrets to improve Defendant's software. (Id. ¶¶ 2, 6-7).
1. Hedge Funds
A hedge fund like Cent Capital is a limited partnership of private investors whose money is managed by professional fund managers who attempt to use investment strategies to earn above-average returns - that is, to beat the market. (See id. ¶¶ 19-20). Beating the market is extremely difficult. Hedge funds must make decisions based on publicly available financial data, which means (in this context) that it is advantageous to quickly process the market data to glean insights that the rest of the world misses. A common source of financial data for hedge funds is Bloomberg, a company that provides this information via its “Bloomberg Terminals,” which can be accessed on a user's computer. (Id. ¶¶ 28, 34). There are over 300,000 of these Bloomberg Terminals in use around the world, and Bloomberg charges a licensing fee for their use. (Id. ¶ 28). While software developers like Wolfram have limited access to Bloomberg data through a developer's license, Cent Capital paid for a full desktop end user license, which is more limited than a developer's license. (Docs. # 164-1 ¶ 49; 178 at 11 & n.4; 164-9 at 15). Hedge funds generally each have their own process for translating raw Bloomberg data into investment decisions. According to Cent Capital, most hedge funds use Excel to execute these calculations. (Doc. # 125 ¶¶ 27, 29). So, in theory, if a hedge fund could develop a tool that consistently analyzed the data more quickly or more insightfully than the use of Excel or other common software permits, that tool could be extremely valuable. (Id. ¶¶ 30-31). Cent Capital claims that it was developing such a tool, which it calls Tabula Rasa. (Id. ¶ 31). Specifically, Cent Capital alleges that Tabula Rasa, once fully built, will be able to analyze financial data more quickly, at a higher volume, and more powerfully than any other software on the market. (Id. ¶ 33).
As Plaintiffs note, “Tabula Rasa” is Latin for “clean slate.” (Doc. # 125 ¶ 33 n.1).
2. Software Programming Background
Wolfram developed the language and computer program that Plaintiffs used to develop a portion of Tabula Rasa. (Docs. # 125 ¶ 23; 164-1 at 2). A software “language” is like other written languages (English, Spanish, or Chinese, for example), in that it provides the equivalent of letters, words, and syntax that allow developers to design software. Wolfram's computer language is called “Wolfram.” (See Doc. # 164-1 at 2). A “function” is a way to accomplish a specific task within one of these languages (like the Excel functions of “SUM” or “AVERAGE”). (Doc. # 164-2 at 14). There are built-in functions in the Wolfram language, but users can also define their own unique functions. (Id.). A software “program” is like an operating system (Windows or Mac, for example) in that it provides some existing structure and tools to allow users to design software. Wolfram's computer program is called Mathematica, and it runs on Wolfram's computer language. (Doc. # 125 ¶ 6). A “platform” is like a program, but in Wolfram's case it describes a more specific set of functions within the Mathematica Program (like Microsoft Word within the Windows operating system). The Wolfram Finance Platform is the relevant platform in this case because it offers tools to analyze and visualize financial data. (See, e.g., Doc. # 164-9 at 32). Wolfram provides “documentation” for every version of its Wolfram Finance Program that explains each built-in function by using tutorials and examples. (Doc. # 164-2 at 17). Also relevant to this case is the Wolfram Enterprise Private Cloud, which is Wolfram's version of a “cloud” that a user can use along with their desktop version of Mathematica (like a Google Document or Microsoft SharePoint). (See Doc. # 125 ¶ 71). To perform operations using the Wolfram Finance Platform, users create “notebooks,” which are plain text files with the extension “.nb” that contain code and metadata related to the code and the notebook. (Doc. # 164-8 at 26). Just as the Wolfram Finance Platform can be likened to Microsoft Word, a notebook can be likened to a Word document. A line of code within these notebooks is called a “cell,” and often has a unique identifier number called a “CellID.” (Id. at 30; Doc. # 164-9 at 20). If a user were to copy and paste a cell from one notebook to another notebook, the CellID would be preserved, even if the content of the cell was later altered. (Docs. # 164-8 at 37; 164-9 at 21).
Cent Capital asserts that while using the Wolfram Finance Platform within Mathematica, it began to develop an early, non-functional version of Tabula Rasa. (Doc. # 125 ¶¶ 77-79). Plaintiffs' goal with Tabula Rasa is to allow users to analyze Bloomberg financial data more rapidly than other investors by utilizing a process that quickly translates a large amount of Bloomberg financial data, data that is not produced in the Wolfram computer language, into the Wolfram language. (Id. ¶¶ 7, 11, 22, 32). The Wolfram language, in turn, purports to allow users to analyze the data in the Wolfram Finance Platform, which offers insightful data analysis. (Id.).
For some time, Plaintiffs attempted on their own to develop Tabula Rasa such that it would perform this function more quickly than anything else available to traders. Plaintiffs' goal was to simultaneously translate the large number of data points that are available through a Bloomberg Terminal full desktop license. Plaintiffs describe this process in terms of “parallel computational kernels.” (Doc. # 138 at 2). In computer software, a “kernel” is a component of an operating system that acts as a messenger between software (like Microsoft Word) and hardware (like computer memory). The court therefore understands the term “running parallel computational kernels” to mean that instead of processing each piece of data one at a time, it was hoped that Tabula Rasa could process data simultaneously, or in “parallel”. (See Doc. # 164-9 at 71-72). Each stock for which Bloomberg provides data has what is called dimensions, which are six data points associated with that stock, including its value at open, high, low, and close, as well as the event count and volume. (See Doc. # 164-3 at 7). Although Plaintiffs have not yet developed a functional version of this code, they theorize that they will eventually be able to process these data dimensions for many stocks simultaneously by way of a connection to Wolfram's cloud services (that is, the Wolfram Enterprise Private Cloud). (Docs. # 125 ¶ 27; 138 at 9; 164-9 at 72). Once these data points are translated, a Tabula Rasa user theoretically could use the Wolfram Finance Platform to perform fact analyses on these data points. (Doc. # 178 at 4).
3. Origins of this Case
This dispute arose after Plaintiffs encountered problems in their own attempts to develop Tabula Rasa. After some time of trying to solve these problems on their own, they contracted with Wolfram to provide a fix. About a month after Wolfram began consulting with Cent under this contract, Wolfram held a webinar (the “January 2021 Webinar”) during which Wolfram distributed certain code to the webinar attendees. (Docs. # 125 ¶ 95; 166 at 2). Cent argues this code contained its trade secrets and that Wolfram had clandestinely incorporated some of Cent's trade secrets into its base software. (Doc. # 125 ¶¶ 2-3, 42, 47, 93-96, 99-100, 102, 110, 119, 128). Although Plaintiffs concede that the individual lines of code and the functions within those lines are not trade secrets (see, e.g., Doc. # 164-9 at 20), Plaintiffs maintain that the lines collectively make up a trade secret because they combine existing techniques in a new way. (Docs. # 125 ¶ 3; 164-9 at 9; 138 at 3; 178 at 8-12). Cent also contends that this contracting relationship fizzled out and that Wolfram therefore never performed what it was contracted to do. (Doc. # 125 ¶¶ 6-7, 11, 47, 64, 68-69, 95, 129). Plaintiffs therefore assert three claims against Defendant: breach of a non-disclosure agreement contract (Count One), breach of a consulting contract (Count Two), and violation of two trade secrets laws (Count Three, under the Defend Trade Secrets Act of 2016, 18 U.S.C. §§ 1836 et seq., and the Illinois Trade Secrets Act, 765 ILCS 1065/1 et seq.).
This appears to be a significant change from what Plaintiffs had previously argued to the court. In an earlier filing, Plaintiffs listed their alleged trade secrets as ten individual lines of code (as well as certain ideas this code represents) that were present in the BloombergTerminal.nb notebook. (Doc. # 138-1 at 1, 16).
Defendant Wolfram counters that: (1) it disclosed only preexisting functionality of its own software at the January 2021 Webinar; and (2) Defendant made no relevant changes to its software after interacting with Plaintiffs, and this second point can be shown by the consistency in its source code before and after the contract with Plaintiffs. (Doc. # 166 at 1).
Plaintiffs respond that although the webinar notebook code “did not appear particularly remarkable, as it was comprised of lines of code which are relatively common in the computer coding world, [] the manner in which those lines were combined and utilized provided webinar attendees with the ‘missing piece' to accomplish what Cent hired Wolfram to help it develop.” (Doc. # 178 at 3) (emphasis in original).
The operative complaint is Plaintiffs' Second Amended Complaint (Doc. # 125). After the parties selected their testifying experts, Defendant moved to bar portions of Plaintiffs' expert testimony (Doc. # 161) and moved for summary judgment (Doc. # 163). This summary judgment motion is now fully briefed (Docs. # 164, 166, 178, 181), and ripe for decision.
The court notes that Plaintiffs' Response (Doc. # 178) fails to fully comply with Appendix II of the court's Initial Order (Doc. # 27). Specifically, Plaintiffs' Response does not directly dispute any of Defendant's statements of undisputed facts, instead asserting legal arguments wherever Plaintiffs dispute those statements. (See, e.g., Doc. # 178 at 8-10). Plaintiffs do not include any proposed statement of undisputed facts, and do not attach any copies of evidentiary material in their Response. The court notes that even though it has reviewed this record, the Eleventh Circuit has held that a district court need not “parse a summary judgment record to search out facts or evidence not brought to the court's attention.” Atlanta Gas Light Co. v. UGI Utilities, Inc., 463 F.3d 1201, 1208 n.11 (11th Cir. 2006). Nevertheless, and notwithstanding this non-compliant briefing - indeed because of it - the court has parsed the undisputed facts by relying on the Rule 56 record as well as reviewing those portions of the briefing and record where Plaintiffs explicitly indicate that they do not dispute Defendant's statements of undisputed facts.
II. Standard of Review
Under Federal Rule of Civil Procedure 56, summary judgment is proper “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law.” Celotex Corp. v. Catrett, 477 U.S. 317, 322 (1986). The party asking for summary judgment always bears the initial responsibility of informing the court of the basis for its motion and identifying those portions of the pleadings or filings which it believes demonstrate the absence of a genuine issue of material fact. Id. at 323. Once the moving party has met its burden, Rule 56 requires the non-moving party to go beyond the pleadings and - by pointing to affidavits, or depositions, answers to interrogatories, and/or admissions on file -designate specific facts showing that there is a genuine issue for trial. Id. at 324.
The substantive law will identify which facts are material and which are irrelevant. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). All reasonable doubts about the facts and all justifiable inferences are resolved in favor of the non-movant. See Allen v. Bd. of Pub. Educ. for Bibb Cnty., 495 F.3d 1306, 1314 (11th Cir. 2007); Fitzpatrick v. City of Atlanta, 2 F.3d 1112, 1115 (11th Cir. 1993). A dispute is genuine “if the evidence is such that a reasonable jury could return a verdict for the nonmoving party.” Anderson, 477 U.S. at 248. If the evidence is merely colorable, or is not significantly probative, summary judgment may be granted. See id. at 249.
When faced with a “properly supported motion for summary judgment, [the nonmoving party] must come forward with specific factual evidence, presenting more than mere allegations.” Gargiulo v. G.M. Sales, Inc., 131 F.3d 995, 999 (11th Cir. 1997). As Anderson teaches, under Rule 56(c) a plaintiff may not simply rest on her allegations made in the complaint; instead, as the party bearing the burden of proof at trial, she must come forward with at least some evidence to support each element essential to her case at trial. See Anderson, 477 U.S. at 252. “[A] party opposing a properly supported motion for summary judgment ‘may not rest upon the mere allegations or denials of his pleading, but . . . must set forth specific facts showing that there is a genuine issue for trial.'” Id. at 248 (citations omitted).
Summary judgment is mandated “against a party who fails to make a showing sufficient to establish the existence of an element essential to that party's case, and on which that party will bear the burden of proof at trial.” Celotex Corp., 477 U.S. at 322. “Summary judgment may be granted if the non-moving party's evidence is merely colorable or is not significantly probative.” Sawyer v. Sw. Airlines Co., 243 F.Supp.2d 1257, 1262 (D. Kan. 2003) (citing Anderson, 477 U.S. at 250-51).
“[A]t the summary judgment stage the judge's function is not himself to weigh the evidence and determine the truth of the matter but to determine whether there is a genuine issue for trial.” Anderson, 477 U.S. at 249. “Essentially, the inquiry is ‘whether the evidence presents a sufficient disagreement to require submission to the jury or whether it is so one-sided that one party must prevail as a matter of law.'” Sawyer, 243 F.Supp.2d at 1262 (quoting Anderson, 477 U.S. at 251-52); see also LaRoche v. Denny's, Inc., 62 F.Supp.2d 1366, 1371 (S.D. Fla. 1999) (“The law is clear . . . that suspicion, perception, opinion, and belief cannot be used to defeat a motion for summary judgment.”).
III. Factual Background
The court has gleaned the facts set out in this opinion from the parties' submissions and the court's own examination of the evidentiary record. All reasonable doubts about the facts have been resolved in favor of Plaintiffs, the nonmoving parties. See Info. Sys. & Networks Corp. v. City of Atlanta, 281 F.3d 1220, 1224 (11th Cir. 2002). These are the “facts” for summary judgment purposes only. They may not be the actual facts that could be established through live testimony at a trial. See Cox v. Adm'r U.S. Steel & Carnegie Pension Fund, 17 F.3d 1386, 1400 (11th Cir. 1994).
1. Wolfram-Cent Consulting Relationship
In December 2020, Cent Capital entered into a consulting relationship with Wolfram. (Docs. # 125-6; 164-19). As part of this contract, Plaintiffs and Defendant signed a Non Disclosure Agreement (“the Agreement”) and a Technical Consulting Development Support Statement of Work Contract (“consulting contract”). (Docs. # 125-13 at 1-2; 178 at 5). Below, the court discusses the content of these agreements before listing the facts related to the actual consulting relationship.
a. The Agreement
Wolfram Research and Cent Holdings signed the Agreement with an effective date of December 4, 2020. (Doc. # 125-13 at 1-2). The Agreement defines Confidential Information as any “information [the parties] consider confidential and/or proprietary.” (Id. at 1). If a party receives confidential information, the Agreement imposes the following duty:
to protect Confidential Information disclosed to it by the other Party (“Discloser”): (i) if it is clearly and conspicuously marked as ‘confidential' or with a similar designation; (ii) if it is identified by the Discloser as confidential before, during, or promptly after presentation or communication; or (iii) if it is disclosed in a manner in which the Discloser reasonably communicated, or the Recipient should reasonably have understood under the circumstances that the disclosure should be treated as confidential, whether or not the specific designation ‘confidential' or any similar designation is used.(Id.). The Agreement limits this duty as follows: “with respect to Confidential Information that: (i) was known to the Recipient through proper means before receipt from the Discloser; (ii) is or becomes publicly available through no fault of the Recipient; . . . [or] (iv) is independently developed by the Recipient without a breach of this Agreement.” (Id.). The Agreement states that it “shall be governed by and construed under the laws of the State of Illinois, without regard to its rules for conflict of laws.” (Id. at 3).
b. Consulting Contract
In the consulting contract, Defendant agreed to “provide technical consulting Development Support . . . to Cent Holding . . . by assigning one or more designated contractor(s) . . . to work as request[s] [are] made by the Customer during the period of this Statement of Work.” (Docs. # 166 ¶ 37; 164-12 ¶ 9; 164-19 at 1; 178 at 15). The contract states that “[t]his work will be . . . advisory or collaborative in nature and Wolfram will not have responsibility for the accuracy or efficiency of any code provided by Wolfram as part of the [Development Support].” (Docs. # 166 ¶ 38; 164-19 at 1; 178 at 15). Requests for work were required to be made in writing in a specific way, and Wolfram was free to accept or reject such requests for any reason “deemed important.” (Docs. # 166 ¶ 39; 164-19 at 1; 178 at 15). The consulting contract indicates that Wolfram would “undertake commercially reasonable efforts to achieve a successful result for Customer” (Doc. # 164-19 at 2) but disclaims any guarantee that the work would result in “any specific deliverable.” (Id.; Docs. # 166 ¶ 40; 178). As the contract repeats later in its text, “[u]nless otherwise indicated herein, the Services do not include project deliverables or guarantees.” (Doc. # 164-19 at 3). The contract also contains a confidentiality provision that states: “All materials provided by Customer that are marked or otherwise clearly identified as being confidential, and are not otherwise already known to Wolfram or publicly available in some other way, shall be protected by Wolfram as such.” (Id. at 2). This contract does not contain a choice-of-law provision. (Id. at 2-3).
c. Consulting Relationship
Wolfram employee Michael Kelly was assigned to consult with Cent. (Doc. # 164-2 at 21). On December 9, 2020, Defendant quoted Plaintiffs a price of $18,400 in a document entitled “Estimate of Services,” and Plaintiffs paid $10,000 as a prepayment to Defendant. (Doc. # 12510 at 2-3). In forming a project schedule, the parties agreed on goals for four phases of the consultation. (Doc. # 125-3). As part of this process, on November 27, 2020, and December 3, 2020, Cent Capital disclosed code to Michael Kelly in approximately six Wolfram language notebooks, called:
• “CENT0000760 - Bloomberg.nb,”
• “CENT0000761 - CentCorpScopeDesign.nb,”
• “CENT0000762 - Initialization.nb,”
• “CENT0000764 - Interface.nb,”
• “CENT0000765 - Lists.nb,” and
• “CENT0000767 - CentCapitalProjectPlan.nb.”
(Doc. # 164-2 at 19). These documents “describe Cent's desire to develop a custom interface for a chart that would display various types of financial data using the preexisting built-in functions and features of Wolfram's software and the Wolfram language.” (Id.). They also “include sample code written [by Cent] in the Wolfram Language” appearing to be “a prototype or an illustrative mockup of the desired interface.” (Id.). The built-in functions that the code uses to access Bloomberg Terminal data are from a package called “BloombergLink,” including the function “MarketDataGet.” (Id. at 19-20).
As part of this consulting process, Wolfram created for Cent Capital a Wolfram language notebook, called “BloombergTerminal.nb.” (Docs. # 125-19; 164-2 at 20). Wolfram sent to Cent Capital roughly four versions of this notebook on the following dates: December 29, 2020, December 31, 2020, February 3, 2021, and February 10, 2021. (Doc. # 164-2 at 21). There are two emails in the record documenting the consulting relationship. The first is on December 29, 2020, from Michael Kelly to Scott Gillaspie (the founder and CEO of Cent Holdings and Chief Investment Officer at Cent Capital). (See Doc. # 125-16). That email attaches and discusses “notebooks that were used in discussion yesterday” and summarizes that the parties covered the following topics: “OHLCV [open, high, low, close, volume] time series” data, linking functions to Bloomberg Terminal, and the need for helper functions to transform nested associations. (Doc. # 125-16). The second email sent on December 31, 2020, again from Kelly to Gillaspie, it summarizes that the parties covered the following topics: “FinancialData,” “TickData,” and “IntradayTickData.” (Doc. # 125-17). This consulting relationship never produced a fully functional product and did not complete any goals listed for the initial or later phases of the consulting relationship. (Docs. # 166 ¶ 29; 178 at 13; 164-9 at 62; 164-8 at 20).
2. Use of Trade Secrets to Improve Defendant's Software
The parties certainly have a number of disagreements in this case. But, the following is not one of them: It is undisputed that Defendant did not incorporate any of Plaintiffs' trade secrets into their products. (Docs. # 178 at 8; 164-9 at 52).
3. The Alleged Trade Secret Dissemination at the January 2021 Webinar
Defendant held a webinar on January 27, 2021, entitled “Financial Data Retrieval with Forecasting and Analyzing Stock Prices,” in which it pitched a product that would allow users to communicate with Bloomberg in the Wolfram language. (Docs. # 164-2 at 4; 166 ¶ 9; 178 at 810). In 2012, long before this webinar took place, Wolfram had already disclosed that its software could be used to analyze “live, trading quality” Bloomberg Terminal data. (Docs. # 166 ¶ 34; 178 ¶ 34; 164-18 at 2). During the January 2021 Webinar, Wolfram offered attendees a free download of a Wolfram language notebook containing this product and displayed this notebook on the screen. (Docs. # 166 ¶ 8; 178 at 8; 164-9 at 7). This notebook was labeled “WOL00006 -Financial Data Retrieval with Forecasting and Analyzing Stock Prices.nb.” (Doc. # 164-2 at 7).
a. Thirteen Lines of Code
The webinar notebook contained many lines of code, but only thirteen of those lines relate to communicating with the Bloomberg Terminal. Therefore, only these thirteen lines are relevant to Plaintiffs' trade secret claim. (Docs. # 166 ¶ 9; 164-3 at 10; 178 at 8). These thirteen lines of code are labeled with headings like “Create a new connection” and “Get Reference Data,” and are “separate, one-line examples showing how Wolfram's preexisting built-in functions can be used.” (Docs. # 164-3 at 2-3; 164-2 at 21 n.2; 178).
None of the code that Cent provided to Wolfram was identical to the webinar notebook code. (Doc. # 164-2 at 4). The code that Cent provided to Wolfram used different functions (“BloombergLink”) to access BloombergTerminal data than the functions (BloombergTerminal service connection) that the webinar notebook used or that Wolfram discussed in the January 21 Webinar. (Docs. # 164-2 at 4, 20; 164-1 at 2). Additionally, none of the thirteen lines individually constitutes a trade secret. (Docs. # 178 at 10; 164-9 at 20, 29-32, 48).
Six of the webinar notebook lines - lines 1, 3, 4, 5, 6, and 7 - match or resemble the BloombergTerminal.nb code that Defendant provided to Plaintiffs as part of the consulting contract. (Docs. # 164-3 at 10; 166 ¶ 10; 164-9 at 20, 70). However, all six of these lines are also present in Wolfram's preexisting documentation (five appear with minor differences such as choosing different options). (Docs. # 164-3 at 10; 166 ¶ 9; 178 at 8-10). Four of these lines -lines 1, 4, 5, and 6 - were copied directly from the BloombergTerminal.nb notebook, as is apparent from their matching CellIDs. (Docs. # 164-9 at 20, 70; 166 ¶ 12). When Wolfram copied these four lines into the webinar notebook, Wolfram did not alter their sequence. (Docs. # 88-27 at 10, 23, 28, 57; 164-3 at 11; 164-9 at 29).
The remaining seven lines of the webinar product code - lines 2, 8, 9, 10, 11, 12, and 13 - do not match or resemble the BloombergTerminal.nb code. (Doc. # 164-3 at 10). Indeed, five of those lines - lines 2, 8, 9, 12, and 13 - are “substantially similar to lines Wolfram used in previous documentation.” (Doc. # 178 at 8; see also Doc. # 164-3 at 10).
Two of the seven lines - lines 10 and 11 - have their functions documented in Wolfram's preexisting documentation. (Doc. 164-3 at 10). Like the other eleven lines, neither of these two lines on their own constitute a trade secret. (Docs. # 178 at 7; 164-9 at 19, 22-23, 28-30, 47). The logic contained within these two lines is also not a trade secret. (Doc. # 178 at 7). The logic of the two lines (10 and 11) is to drop the value “EventCount” from the Bloomberg data so Wolfram's function “TradingChart” can process this data. (Doc. # 164-9 at 45).
b. Trade Secret of Code Logic
Plaintiffs' alleged trade secrets can be summarized as follows: (1) transforming data received from the Bloomberg Terminal service so that it can be displayed in Mathematica; (2) analyzing the data in “real time” for many financial instruments with multiple processes running at the same time; and (3) displaying the data in a customized way. (Docs. # 164-9 at 66, 48-49; 166 ¶ 30; 178 at 13). The court examines each of these categories below.
i. Transforming Data
The idea of transforming data from Bloomberg's to Wolfram's language is documented as early as May 15, 2012, when Wolfram announced the release of its Finance Platform. (Doc. # 164-18 at 2). Several Wolfram functions transform Bloomberg data into the Wolfram Language, and those functions are identified in documentation that predates the Wolfram-Cent relationship. (Docs. # 164-1 ¶ 4; 164-13; 164-14).
The only type of data transformation identified in the webinar notebook is the removal of the “EventCount” field from the Bloomberg data for use in a “TradingChart” function. (Docs. # 166 ¶ 31; 164-9 at 65; see also Doc. # 164-9 at 36-37). Wolfram's preexisting documentation explains the need for this type of transformation, as it discusses “(1) the format of the data received as output from EntityValue; (2) the format of the data needed as input into TradingChart; and (3) numerous built-in functions for rearranging data from one format to another.” (Doc. # 164-2 at 27; see also Docs. # 164-2 at 5-6; 164-15 at 1; 164-12 ¶ 3). Such “selection and manipulation of data” is “an essential skill for anyone writing a program” (Doc. # 164-2 at 25) and having “an array of data values that nee[d] to be transformed” is “a very common problem in data analysis.” (Doc. # 164-9 at 32).
The webinar notebook and the BloombergTerminal notebook both instruct Mathematica to drop “EventCount” data, but the webinar notebook uses the function “EntityValue” to do this, while the BloombergTerminal.nb notebook uses the function “ServiceExecute.” (Doc. # 164-2 at 27).
ii. Analyzing Data in “Real Time”
There is no serious dispute in this case that the concept of analyzing data in “real time,” i.e., by running processes at the same time, was not explicitly disclosed at the January 2021 Webinar. (Docs. # 166 ¶ 32; 178 at 13).
iii. Displaying Data in Customized Way
Displaying the data in a customized way was not disclosed explicitly or implicitly at the January 2021 Webinar. (Docs. # 166 ¶ 33; 178 at 14). The only functions that were displayed and disseminated in the webinar were Wolfram's own built-in functions. (Doc. # 166 ¶ 33; 178 at 14; 164-3 at 17, 21, 23).
In its legal analysis, the court considers the various arguments raised by the parties in their respective briefs.
IV. Analysis
As noted above, Plaintiffs assert the following claims against Defendant: breach of the Non-Disclosure Agreement (Count One), breach of the consulting contract (Count Two), and violation of two trade secrets laws (Count Three, listing the Defend Trade Secrets Act of 2016, 18 U.S.C. §§ 1836 et seq., and the Illinois Trade Secrets Act, 765 ILCS 1065/1 et seq.). The court analyzes each count in the order alleged.
1. Breach of Non-Disclosure Agreement (Count One)
The fact section of the Non-Disclosure Agreement contains a choice-of-law clause that specifies Illinois law as the governing law of the contract. Neither party disputes that Illinois law applies under this choice-of-law provision (see Docs. # 125, 166, 178, 181), and so the court applies Illinois law to Plaintiffs' breach of contract claim.
Under Illinois law, to establish a breach of contract a plaintiff must show: “(1) the existence of a valid and enforceable contract; (2) performance by the plaintiff; (3) breach of the contract by the defendant; and (4) resultant injury to the plaintiff.” Gonzalzles v. American Exp. Credit Corp., 733 N.E.2d 345, 351 (Ill.App.Ct. 2000). Here, neither party contests that the NonDisclosure Agreement was a valid and enforceable contract. (Cf. Docs. # 125 ¶¶ 1, 4, 89, 102; 166; 178 at 5; 181). Additionally, neither party contests that Plaintiffs performed under the contract and did not breach their duties to not disclose any confidential information that Defendant provided to Plaintiffs. The main contract dispute between the parties relates to the third and fourth prongs.
Regarding the third prong, the question is whether Defendant breached the contract. In relevant part, the Non-Disclosure Agreement imposes a duty on the recipient of confidential information not to disclose it if the discloser of such information “clearly and conspicuously” marks it as confidential or identifies it as confidential around the time the information is communicated. (Doc. # 125-13 at 1).
Because it is undisputed that the information that Defendant disclosed during the January 2021 Webinar was information that Defendant sent to Plaintiffs, it appears that Defendant was the discloser of the information that Plaintiffs assert is confidential information. Even considering Plaintiffs' argument that they disclosed certain ideas to Defendant as part of their consulting relationship, and thus that they could be the disclosers (See Doc. # 178 at 7), the only disclosure at issue here relates to the application of the Agreement with respect to the January 2021 Webinar. (Doc. # 125 ¶¶ 2, 102). But, Plaintiffs face an insurmountable hurdle in advancing this claim. It is undisputed that none of the code that Cent provided to Wolfram was identical to the webinar notebook code. (Doc. # 164-2 at 4). Additionally, the code that Cent provided to Wolfram used different functions (“BloombergLink”) to access BloombergTerminal data than the functions (BloombergTerminal service connection) than those the webinar notebook used or that Wolfram discussed in the January 21 Webinar. (Docs. # 164-2 at 4, 20; 164-1 at 2).
As Plaintiffs have not presented sufficient evidence that Defendant breached the Agreement by disclosing Plaintiffs' confidential information, the court need not analyze the fourth prong (which requires they show they suffered an injury).
It is undisputed that Defendant did not disclose at the January 2021 Webinar any code that Plaintiffs provided to Defendant. Therefore, Defendant is entitled to summary judgment on the claim that it breached the Agreement.
2. Breach of Consulting Contract (Count Two)
Plaintiffs claim that Defendant “breached the Consulting Contract by (1) failing to protect Plaintiffs' confidential information, (2) failing to deliver the ‘milestones/tasks' as agreed, and (3) ceasing to perform work.” (Doc. # 125 ¶ 119).
Initially, the court must consider which state's law applies to the interpretation of the parties' consulting contract. Unlike the Agreement, which contained an express choice-of-law provision, the consulting contract is silent on this question. (Doc. # 125-6 at 2-3). As a federal court sitting in diversity, this court applies the principles of conflict of laws of the state in which it sits. Here, that is Alabama. See O'Neal v. Kennamer, 958 F.2d 1044, 1046 (11th Cir. 1992). “Alabama applies the traditional doctrines of lex loci contractus to contract claims.” Colonial Life & Acc. Ins. Co. v. Hartford Fire Ins. Co., 358 F.3d 1306, 1308 (11th Cir. 2004). This doctrine holds that “a contract is governed by the laws of the state where it is made except where the parties have legally contracted with reference to the laws of another jurisdiction.” Cherry, Bekaert & Holland v. Brown, 582 So.2d 502, 506 (Ala. 1991). Defendant has asserted that they “assum[e] Alabama law applies” (Doc. # 166 at 32 n.5), as they believe that the contract was made in the state of Alabama. Plaintiffs have not offered a position on this question. (Cf. Doc. # 178). The contract contains signatures from both parties. One lists an address in Illinois and the other lists an address in Alabama. (Doc. # 125-6 at 3). Because there does not appear to be a dispute over applying Alabama law, and because it appears this contract was made in Alabama, the court applies Alabama law to this breach of contract claim. But, this analysis is largely academic because as relevant here the legal elements for breach of contract in Illinois and Alabama are nearly identical. Compare Gonzalzles v. American Exp. Credit Corp., 733 N.E.2d 345, 351 (Ill.App.Ct. 2000) with Southern Medical Health Syst., Inc. v. Vaughn, 669 So.2d 98, 99 (Ala. 1995).
Under Alabama law, a breach of contract claim requires a plaintiff to prove “(1) the existence of a valid contract binding the parties in the action, (2) his own performance under the contract, (3) the defendant's nonperformance, and (4) damages.” Vaughn, 669 So.2d at 99. As with the Agreement analyzed above, there is no dispute that this consulting agreement was a valid contract binding the parties, nor is it contested that Plaintiffs performed their end of the contract (for example, by paying Defendant a fee). The main dispute arises with respect to the third and fourth elements. Below, the court considers Plaintiffs' theories and explains why the undisputed facts in this case show that Plaintiffs have failed to present sufficient Rule 56 evidence to support the third prong of their breach of contract action.
First, the court considers the claim that Defendant failed to protect Plaintiffs' confidential information. In relevant part, the consulting contract contains a paragraph marked “Confidentiality” that states: “All materials provided by Customer that are marked or otherwise clearly identified as being confidential, and are not otherwise already known to Wolfram or publicly available in some other way, shall be protected by Wolfram as such.” (Doc. # 125-6 at 2).
Here, the “Customer” is clearly Cent. And it is undisputed that Cent provided to Wolfram six notebooks containing what Plaintiffs allege are trade secrets. (Docs. # 164-2 at 19; 125 ¶ 81). However, it is also undisputed that none of the code that Cent provided to Wolfram was identical to the webinar notebook code. (Docs. # 164-2 at 4; 164-3 at 10). Further, it is undisputed that Cent did not provide any code to Wolfram that related to accessing Bloomberg Terminal data, which was the subject of the relevant portion of the January 2021 Webinar notebook. (Docs. # 164-2 at 4; 178 at 9-10). Because Defendant did not fail to protect any of the information that Plaintiffs provided in the six notebooks, Defendant is entitled to summary judgment on this portion of this claim.
Plaintiffs also argue that Defendant violated the confidentiality provision of this agreement by copying “lines of code directly from a notebook specifically produced as part of its relationship with Cent and pasted those cells into a notebook distributed to attendees of its January 2021 Webinar.” (Doc. # 178 at 21) (emphasis in original). While it is undisputed that Defendant copied four lines from the BloombergTerminal.nb notebook into the January 2021 Webinar notebook, this does not establish a violation of the confidentiality provision of the consulting contract. That confidentiality provision merely states that “All materials provided by Customer that are marked or otherwise clearly identified as being confidential, and are not otherwise already known to Wolfram or publicly available in some other way, shall be protected by Wolfram as such.” (Doc. # 125-6 at 2) (emphasis added). It is undisputed that Defendant provided the BloombergTerminal.nb notebook to Plaintiffs. This notebook was not part of the “materials provided by Customer,” such that it would fall under the protection of the confidentiality provision of the consulting agreement. Therefore, Plaintiffs' arguments as to the breach of the confidentiality provision fail, and summary judgment is due to be entered in favor of Defendant on this portion of the claim as well.
Second, Plaintiffs argue that Defendant failed “to deliver the ‘milestones/tasks' as agreed.” (Doc. # 125 ¶ 119). Plaintiffs' Second Amended Complaint also alleges that “Wolfram's deliverable work (the ‘milestones/tasks') under the consulting [contract] has been slower than anticipated, and now Wolfram has ceased (or appears to have ceased) performing any meaningful work at all.” (Id. ¶ 116). Defendant's response to this argument is that “[t]he unambiguous contract language [] disclaims ‘any specific deliverable' and otherwise limits Wolfram's obligations justifies summary judgment in Wolfram's favor.” (Doc. # 166 at 33). Plaintiffs respond that this was “in part due to the misfeasance on the part of Wolfram.” (Doc. # 178 at 10). In its Reply, Defendant has not addressed this last point. (Cf. Doc. # 181).
The contract language stipulates that “Wolfram Research, Inc.... will provide technical consulting Development Support . . . to Cent Holding . . . by assigning one or more designated contractor(s) . . . to work as request[s] [are] made by the Customer during the period of this Statement of Work.” (Doc. # 125-6 at 2). The agreement further states that “[t]his work will be [] advisory or collaborative in nature and Wolfram will not have responsibility for the accuracy or efficiency of any code provided by Wolfram as part of the [Development Support].” (Id.). Requests for work were required to be made in writing in a specific way, and Wolfram was free to accept or reject such requests for any reason “deemed important.” (Id.). The consulting agreement indicates that Wolfram “will undertake commercially reasonable efforts to achieve a successful result for Customer” (id.) but expresses that “Wolfram makes no Warranty that the Services shall result in any specific deliverable.” (Id.). The agreement repeats later that “[u]nless otherwise indicated herein, the Services do not include project deliverables or guarantees.” (Id. at 3).
In forming a project schedule, the parties agreed on goals related to four phases of the consultation. (Doc. # 125-3). Wolfram created for Cent Capital a Wolfram language notebook, called “BloombergTerminal.nb,” as part of this consulting process. (Docs. # 125-19; 164-2 at 20). This consulting relationship did not produce a fully functional product and did not fully complete any phase of deliverables. (Docs. # 166 ¶ 29; 178 at 13; 164-9 at 62; 164-8 at 20).
Given each of these undisputed facts, the question for the court is whether Defendant's performance under the contract breached the contract's terms. Under the terms of the contract, the only thing that Wolfram promised to do was to “provide technical consulting Development Support . . . to Cent Holding . . . by assigning one or more designated contractor(s) . . . to work as request[s] [are] made by the Customer during the period of this Statement of Work.” (Doc. # 125-6 at 2). Wolfram did this by assigning Michael Kelly as a “designated contractor[].” (Id.). Otherwise, the contract expressly warrants against “any specific deliverable” (id.), and even repeats this on the next page, warning “[u]nless otherwise indicated herein, the Services do not include project deliverables or guarantees.” (Id. at 3). The plain language of the consulting contract therefore requires only that Wolfram provide technical consulting development support to Cent by assigning a designated contractor to work as “request[s] [are] made.” (Doc. # 125-6 at 2). And here, it is undisputed that Wolfram did just that. Although the consulting relationship did not produce a fully functional product, the agreement twice warns that Wolfram does not have the duty to produce any functional product.
Further, to Plaintiffs' argument that Wolfram has ceased (or appears to have ceased) performing any meaningful work at all” (Doc. # 125 ¶ 116), nothing in the contract requires Wolfram to continue performing any work. As the contract notes that “Wolfram reserves the right to accept or reject such request based on available resources or any other reason deemed important.” (Doc. # 125-6 at 2) (emphasis added). Once again, the plain language of the contract allows Wolfram to reject Cent's work requests for any reason it deems important. In addition to the presence of this language, the absence of any language requiring Wolfram to continue performing meaningful work indicates that the contract simply does not require Wolfram to do this. Wolfram performed its duties under this contract by assigning a designated contractor to work as requests were made, and by not failing to protect any materials provided by Cent. Defendant is entitled to summary judgment on Plaintiffs' claim in Count Two.
3. Breach of Trade Secrets Laws (Count Three)
Plaintiffs next claim is that Defendant violated the Defend Trade Secrets Act of 2016 (“DTSA”), 18 U.S.C. §§ 1836 et seq., and the Illinois Trade Secrets Act (“ITSA”), 765 ILCS 1065/1 et seq. (Doc. # 125 ¶ 122). They allege that their “tool known as Tabula Rasa and Plaintiffs' process for implementing Tabula Rasa on a Bloomberg Terminal with the Wolfram language are ‘Trade Secrets' within the meaning of the Defend Trade Secrets Act . . . and the Illinois Trade Secrets Act.” (Id.). Plaintiffs allege that once they provided Defendant with access to these secrets (id. ¶ 127), Defendant misappropriated them by disclosing them in “at least one webinar” (id. ¶ 130) and by using them “to create improvements and/or modifications to Wolfram's software.” (Id. ¶ 129).
In Defendant's Motion for Summary Judgment, it argues that the experts now agree that Defendant did not use any trade secrets of Plaintiffs to improve or modify Defendant's software. (Doc. # 166 at 19). As discussed above, this is now undisputed. (See Doc. # 178 at 8, 16). Therefore, Defendant is due summary judgment related to this theory.
As to Plaintiffs' other theory, that Defendant's disclosure of their trade secrets in the January 2021 Webinar violated federal and state trade secrets laws, Defendant highlights that none of the individual lines of the January 2021 Webinar notebook contained any trade secrets or confidential information. (Doc. # 166 at 21). Defendant also addresses Plaintiffs' argument that the combination of the lines in the January 2021 Webinar is a dissemination of a trade secret. As Defendant puts it, “[t]his is a hypothetical argument that Cent does not and cannot substantiate with evidence.” (Id. at 27). Defendant argues that Plaintiffs have never disclosed what the specific “trade secret” is, and that Plaintiffs have not sufficiently pointed to Rule 56 evidence showing that any such secret can meet other requirements (such as having economic value, being owned by Plaintiffs, and being kept secret by Plaintiffs). (Id. at 28-31). Finally, Defendant argues that even if there were such a trade secret combination, there is no evidence that it was disclosed at the January 2021 Webinar. (Id. at 31).
Plaintiffs respond that although no individual line of code in the webinar notebook is a trade secret, “the novelty of its idea lay in the unique combination of these relatively common elements to create a previously unexplored capability.” (Id. at 18).
Defendant's Reply argues that it is “impossible” for Plaintiffs to contend that the logic of the entire webinar notebook discloses a trade secret because the webinar and BloombergTerminal.nb notebooks are “completely different.” (Id. at 5). Below, the court considers these arguments and the summary judgment facts on which they are based. After careful review, the court concludes that there is no genuine dispute of material fact as to this claim.
a. Choice of Law
Initially, the court must consider whether the relevant state trade secrets law is that of Illinois or Alabama. Although Plaintiffs assert a claim under the ITSA, Defendant argues that the Alabama Trade Secret Act should apply because Alabama is where Plaintiffs suffered the alleged economic impact. (Doc. # 166 at 27 n.4). Again, as a federal court sitting in diversity, this court applies the principles of conflict of laws of the state in which it sits - namely, Alabama. See O'Neal, 958 F.2d at 1046. Alabama's relevant conflict of laws principles sound in tort law. The Alabama Supreme Court has acknowledged that the Alabama Trade Secrets Act “replac[ed] common-law tort remedies for the misappropriation of trade secrets.” Allied Supply Co., Inc. v. Brown, 585 So.2d 33, 37 (Ala. 1991). And in tort actions, Alabama uses the lex loci delicti rule from the Second Restatement of Torts. Fitts v. Minnesota Min. & Mfg. Co., 581 So.2d 819, 823 (Ala. 1991). In other words, Alabama applies the law of the “site of the injury, or the site of the event that created the right to sue.” Glass v. Southern Wrecker Sales, 990 F.Supp. 1344 (M.D. Ala. 1998) (quoted in Ex Parte U.S. Bank Nat'l Ass'n, 148 So.3d 1060, 1071 (Ala. 2014)).
Applying these principles to the case at hand, all the members of Cent Holding and Cent Capital are residents and citizens of the state of Alabama. (Doc. # 125 ¶¶ 12-13). Although these two companies are formed under the laws of Delaware, its members reside in Alabama and therefore, any economic impact from alleged harm would likely have occurred in Alabama. In any event, there is no evidence in this record that indicates Plaintiffs would have suffered economic harm in Illinois; so the ITSA simply is not the state trade secrets law that applies here. Because it is likely that if Plaintiffs suffered any economic harm, that harm would have occurred in Alabama, Alabama state law applies to Plaintiffs' state law claims. So, the court construes Plaintiffs' trade secret claim to be advanced under the Alabama Trade Secrets Act (“ATSA”), Ala. Code § 8-27-1 et seq.
The court also notes that the relevant statutory language is similar between the ATSA and the ITSA. Compare Ala. Code § 8-27-1 et seq. with 765 ILCS 1065/1 et seq.
b. Rule of Law
The ATSA defines trade secret as follows:
information that:
a. Is used or intended for use in a trade or business;
b. Is included or embodied in a formula, pattern, compilation, computer software, drawing, device, method, technique, or process;
c. Is not publicly known and is not generally known in the trade or business of the person asserting that it is a trade secret;
d. Cannot be readily ascertained or derived from publicly available information;
e. Is the subject of efforts that are reasonable under the circumstances to maintain its secrecy; and f. Has significant economic value.
Similarly, the DTSA protects:
all forms and types of financial, business, scientific, technical, economic, or engineering information, including patterns, plans, compilations, program devices, formulas, designs, prototypes, methods, techniques, processes, procedures, programs, or codes, whether tangible or intangible, and whether or how stored, compiled, or memorialized physically, electronically, graphically, photographically, or in writing if - (A) the owner thereof has taken reasonable measures to keep such information secret; and (B) the information derives independent economic value, actual or potential, from not being generally known to, and not being readily ascertainable through proper means by, another person who can obtain economic value from the disclosure or use of the information[.]
18 U.S.C. § 1839(3). Because the language of the two statutes is “nearly identical,” this court discusses the statutes together. See Parker v. Petrovics, 2020 WL 3972761, at *4 (N.D. Ala. July 14, 2020).
To survive summary judgment on both their ATSA and DTSA claims, Plaintiffs must establish that there is a material dispute of fact as to: (1) whether Plaintiffs have a trade secret, and (2) whether Defendant disclosed that secret (or those secrets) in the January 2021 Webinar. Although courts have sometimes incanted that the question of “[w]hether something is a trade secret is [one] typically ‘resolved by a fact finder after full presentation of evidence from each side,'” Parker, 2020 WL 3972761, at *6 (quoting Compulife Software Inc. v. Newman, 959 F.3d 1288, 1311 (11th Cir. 2020)), a court may resolve such a claim at the summary judgment stage if there is no genuine dispute of fact as to whether it is possible for Plaintiffs to prove a necessary element of the relevant trade secrets laws. Cf. Movement Mortgage, LLC v. CIS Financial Servs., 670 F.Supp.3d 1282, 1287 (N.D. Ala. 2023) (allowing a jury to decide whether something was a trade secret only because a plaintiff described “the distinctive features of the [secret] with sufficient particularity”); Yellowfin Yachts, Inc. v. Barker Boatworks, LLC, 898 F.3d 1279, 129899 (11th Cir. 2018) (affirming grant of summary judgment finding “no reasonable jury could find that [something] constituted a trade secret,” id. at 1299).
c. Application of Law
The court structures its analysis in relation to Plaintiffs' allegations about their trade secrets claims. So, the court addresses these matters: (1) transforming data received from the Bloomberg Terminal service so that it can be displayed in Mathematica; (2) analyzing the data in “real time” for many financial instruments with multiple processes running at the same time; and (3) displaying the data in a customized way. (Docs. # 164-9 at 66, 48-49; 166 ¶ 30; 178 at 13).
i. Plaintiffs' First Purported Trade Secret
As to the first alleged trade secret, the two questions that remain are whether the secret that Plaintiffs have described in their briefing fits the legal definition of a trade secret, and if so, whether that secret was disclosed in the webinar notebook.
The following points are undisputed: (1) the idea of transforming data from Bloomberg's to Wolfram's language is not a trade secret (Doc. # 164-1 ¶ 4); (2) no individual line of code in the webinar notebook discloses a trade secret (Doc. # 178 at 6-8); and (3) the internal logic of the data transformation present in the webinar notebook is not a trade secret. (Doc. # 178 at 7). Nevertheless, Plaintiffs maintain that the combination, organization, and sequencing of the lines of code implementing this data transformation establishes this concept as a trade secret. (Doc. # 178 at 9). Defendant characterizes this as a “disclosure of a combination” theory. (Doc. # 166 at 28). However that point may be characterized or nominated, for the reasons discussed further below, Plaintiffs have failed to specifically describe their “combination” trade secret, and further have not pointed to Rule 56 evidence that supports their claims.
To start, the fact that the BloombergTerminal.nb notebook contains publicly available code does not doom Plaintiffs' trade secrets claims. Movement Mortgage, LLC v. CIS Financial Servs., Inc., 670 F.Supp.3d 1282, 1287 (N.D. Ala. 2023) (noting this when discussing the ATSA). What Plaintiffs must show to survive summary judgment is that there is a material dispute of fact as to whether the BloombergTerminal.nb contains “a unique combination of that information, which adds value to the information.” Penalty Kick Mgmt. Ltd. v. Coca Cola Co., 318 F.3d 1284, 1291 (11th Cir. 2003) (discussing Georgia trade secrets law) (cited in Movement Mortgage, LLC, 670 F.Supp.3d at 1287). The parties have not cited any binding authority that interprets the requirements for proving a combination trade secrets theory under the DTSA or the ATSA. (See Docs. # 166 at 27-29 (citing cases from the Northern District of Illinois, Northern District of Georgia, Second Circuit, and Eastern District of New York); 178 (citing no cases on this point); 181 at 9 (citing cases from the same jurisdictions)). Further, based on its own research, the court is unaware of any binding authority interpreting the requirements for proving a combination theory under the DTSA or the ATSA. See, e.g., Penalty Kick Mgmt. Ltd., 318 F.3d at 1291 (interpreting Georgia trade secrets law); Capital Asset Research Corp. v. Finnegan, 160 F.3d 683, 688 (11th Cir. 1998) (same).
The court therefore considers the general requirements for specificity in pleading any trade secret under the DTSA or ATSA and applies those to this Rule 56 context. The Third Restatement of Unfair Competition requires a plaintiff to define “the information for which protection is sought with sufficient definiteness to permit a court to apply the criteria for protection . . . and to determine the fact of an appropriation.” Restatement (Third) of Unfair Competition § 39 cmt. d (Am. L. Inst. May 2023 update). The Eleventh Circuit has interpreted the specificity requirements of the ATSA, reversing and remanding for a new trial a jury verdict that found that certain processes were trade secrets because the plaintiff's experts did not specifically identify what the trade secrets were. Peat, Inc. v. Vanguard Research, Inc., 378 F.3d 1154, 1165 (11th Cir. 2004). In that case, the plaintiff's experts defined the trade secret as the “way the whole system fits together,” id. at 1163, asserted that “the trade secret is where it's placed in the process,” id. at 1164, and concluded “a lot of these are tied together. Its not like you can just separate them. They are kind of in groups.” Id. (alteration in original). The court also finds persuasive a Sixth Circuit opinion interpreting the specificity requirements for trade secrets plaintiffs. They must identify the purported trade secret with “reasonable particularity.” Caudill Seed & Warehouse Co., Inc. v. Jarrow Formulas, Inc., 53 F.4th 368, 381 (6th Cir. 2022) (interpreting Kentucky trade secrets law). The Caudill court further noted that “a plaintiff asserting a combination trade secret over highly complex technical information cannot merely offer lists of broad technical concepts identifying categories of information without showing which information contained within those categories constituted a trade secret.” Id. (cleaned up and quotation omitted).
Under these principles, Plaintiffs have not defined their combination trade secret with sufficient specificity for purposes of Rule 56. Plaintiffs accuse Defendant of disclosing “the missing piece to accomplish what Cent contracted them to assist with” (Doc. # 178 at 18-19), and claim that what Cent wanted to create - a “new portfolio construction technique” - required development of “new computer logic, embodied as computer code written in the Mathematica computer language, that would permit Cent (and anyone licensed to do so by Cent) to convert data obtained from Bloomberg using a full desktop license, purchased from Bloomberg, into a format that would allow for the use of that data in Mathematica.” (Id. at 4). Further, Plaintiffs describe their combination trade secret in vague terms, such as “us[ing] an off-the-shelf software product, such as Wolfram's Mathematica, and create[ing] a new method for real-time dynamic cross-market portfolio construction” (id.), putting code in “a certain sequence constructed in a novel way to develop and bring to usefulness a novel and innovative idea” (id. at 10), and “the entire notebook and how the lines of Mathematica code it contains Wolfram under contract with Cent assembled in precisely the right way in exactly the correct sequence to bring about the usefulness that Cent sought.” (Id. at 12) (emphasis in original). Plaintiffs' expert adds little or no specificity, describing the trade secrets as “code, to permit the conversion of Bloomberg data, in the format provided with the full Bloomberg Desktop License, into a form that allowed using that data with Mathematica” (Doc. # 164-8 at 6), and “[t]he steps necessary to take data from the Bloomberg Terminal to run it through these computational mechanisms to modify the data or to transform the data.” (Doc. # 164-9 at 49).
Although Plaintiffs refer to technical terms like “new computer logic, embodied as computer code” (Doc. # 178 at 4) or “a new method for real-time dynamic cross-market portfolio construction” (id.), they have yet to specify how this “logic,” “code,” or “method” embodies a trade secret through its combination, organization, or sequencing. And, their trade secret does not appear to be the overall idea that Plaintiffs describe, as Plaintiffs' expert concedes that “[c]onstructing a platform that can perform real-time equity portfolio construction is not a new idea.” (Doc. # 164-8 at 19) (emphasis in original). But instead of detailing what makes this method unique, or even what the method consists of, Plaintiffs vaguely assert that their trade secret is “the entire notebook and how the lines of Mathematica code it contains Wolfram under contract with Cent assembled in precisely the right way in exactly the correct sequence to bring about the usefulness that Cent sought.” (Doc. # 178 at 12) (emphasis in original). Instead of specifying the exact trade secret that they claim Defendant disseminated, Plaintiffs are doing the very thing the Caudill court warned against: “offer[ing] lists of broad technical concepts identifying categories of information without showing which information contained within those categories constituted a trade secret.” Caudill, 53 F.4th at 381 (cleaned up and quotation omitted).
At the summary judgment stage, the party asking for summary judgment always bears the initial responsibility of informing the court of the basis for its motion and identifying those portions of the pleadings or filings which it believes demonstrate the absence of a genuine issue of material fact. Celotex Corp. v. Catrett, 477 U.S. 317, 323 (1986). Defendant (the movant) has done this by highlighting Plaintiffs' lack of specificity as to what their trade secret is. (Doc. # 166 at 29). Once the moving party has met its burden, Rule 56 requires the non-moving party (here, Plaintiffs) to go beyond the pleadings and - by pointing to affidavits, or depositions, answers to interrogatories, and/or admissions on file - designate specific facts showing that there is a genuine issue for trial. Celotex Corp., 477 U.S. at 324. Under Rule 56(c), a plaintiff may not simply rest on his allegations made in the complaint; instead, as the party bearing the burden of proof at trial, he must come forward with at least some evidence to support each element essential to his case at trial. See Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 252 (1986). But even resolving all reasonable doubts about the facts and all justifiable inferences in favor of Plaintiffs, see Allen v. Bd. of Pub. Educ. for Bibb Cnty., 495 F.3d 1306, 1314 (11th Cir. 2007); Fitzpatrick v. City of Atlanta, 2 F.3d 1112, 1115 (11th Cir. 1993), Plaintiffs' evidence fails to support an essential element of proving a trade secret: namely, defining the secret.
The court also agrees with Defendant that there is no dispute of material fact as to whether the alleged trade secret of transforming Bloomberg data had economic value. As the statutory language quoted above indicates, the ATSA and DTSA both require that a secret must have some economic value. Ala. Code § 8-27-2(1) (“significant economic value”); 18 U.S.C. § 1839(3) (“independent economic value”). Critically, Defendant and the court are still unclear as to what Plaintiffs' trade secret is, and this precludes any meaningful analysis of whether the secret has economic value.
Plaintiffs also failed to show a material dispute of fact as to whether their trade secrets were disseminated in the January 2021 Webinar notebook. The only type of data transformation identified in the webinar notebook is the removal of the “EventCount” field from the Bloomberg data for use in a “TradingChart” function. (Doc. # 166 ¶ 31; 164-9 at 65; see also Doc. # 164-9 at 36-37). Wolfram's preexisting documentation explains the need for this type of transformation, discussing “(1) the format of the data received as output from EntityValue; (2) the format of the data needed as input into TradingChart; and (3) numerous built-in functions for rearranging data from one format to another.” (Doc. # 164-2 at 27; see also Docs. # 164-2 at 5-6; 178 at 6-8). And, it is undisputed that such “selection and manipulation of data” is “an essential skill for anyone writing a program” (Doc. # 164-2 at 25) and illustrates “a very common problem in data analysis.” (Doc. # 164-9 at 32). Not only is this a common problem that Wolfram has documented solutions for, but the webinar notebook and the BloombergTerminal notebook solve this problem differently. The webinar notebook uses the function “EntityValue” to drop “EventCount” data, while the BloombergTerminal.nb notebook uses the function “ServiceExecute.” (Doc. # 164-2 at 27).
Even if Plaintiffs could continue to assert that somehow this data transformation was accomplished in a unique way through a combination of lines or functions, Plaintiffs never specifically defined this trade secret. (Doc. # 166 at 28). Therefore, for the same reasons discussed above regarding whether Plaintiffs have a trade secret, Plaintiffs have not presented evidence that identifies a combination trade secret that was disseminated in the January 2021 Webinar notebook.
Reviewing Plaintiffs' strongest evidence in this case does not rehabilitate this claim. Although there is evidence that six lines of webinar code - lines 1, 3, 4, 5, 6, and 7 - match or resemble the BloombergTerminal.nb code (Docs. # 164-3 at 10; 166 ¶ 10; 164-9 at 20, 70), they are all documented either verbatim or with “minor differences” in Wolfram's preexisting, precent documentation. (Doc. # 164-3 at 10). Another piece of evidence that could otherwise be strong for Plaintiffs is that four of these lines (lines 1, 4, 5, and 6 in the webinar notebook) have CellIDs that match the CellIDs of and appear in the same sequence as four lines in the BloombergTerminal.nb. (Doc. # 88-27 at 10, 23, 28, 57; see also Doc. # 164-3 at 11), demonstrating that a Wolfram employee had the two notebooks open at the same time (or close to the same time). (See Doc. # 164-9 at 20). This may be circumstantial evidence that makes it more probable that Defendant copied something. Crucially, however, this evidence does not rehabilitate Plaintiffs' trade secret claim because it does not help Plaintiffs define what specific trade secret the webinar notebook disclosed. Plaintiffs cannot continue to conclusory argue that the combination of these lines within the webinar notebook is somehow a trade secret. That is, Plaintiffs must point to some evidence that shows why, even though each of the lines of code and their functions are publicly available in Wolfram's preexisting documentation, their combination constitutes a trade secret.
As the court has now repeatedly emphasized, at this point in the litigation, Plaintiffs must specifically define what their trade secret is. Discovery has closed. Both sides have hired experts, and those experts have produced reports. (Docs. # 164-2; 164-8). Plaintiffs have submitted two lists of alleged trade secrets (Docs. # 99, 138) and filed three complaints (Docs. # 1, 47, 125). The court held a “Computer Science Day” to allow the parties to clarify the technical concepts in this case. (See Doc. # 116). The two Wolfram notebooks have been identified and the parties (and their experts) have parsed each line of code with fine tooth combs. (See, e.g., Docs. # 164-3; 164-4; 164-5; 164-7; 164-9 at 20, 29-32, 48). Within those notebooks are, at most, thirteen lines of code that Plaintiffs allege contain trade secrets. But, Plaintiffs still have not yet specified what exactly is unique about their combination, organization, or sequencing of code, and they have not highlighted what about that code was a secret that was disseminated in the webinar notebook. Because after multiple opportunities over three years they have not done this, Defendant is entitled to summary judgment as to Plaintiffs' first alleged trade secret.
ii. Plaintiffs' Second Purported Trade Secret
As to Plaintiffs' second “trade secret,” it is undisputed that the January 2021 Webinar did not disclose anything related to analyzing data in “real time” by running processes at the same time. (Docs. # 166 ¶ 32; 178 at 13).
Although Plaintiffs concede this, they argue that this concept was indirectly disclosed because the thirteen lines of code gave attendees a “missing piece” that would ultimately allow them to accomplish what Plaintiffs were trying to do. (Doc. # 178 at 13). Plaintiffs do not elaborate on this “missing piece” argument or why it enabled those attending the webinar to process data differently. Nevertheless, it does not amount to a violation of either the ATSA or the DTSA.
The ATSA sets out that “[a] person who discloses or uses the trade secret of another, without a privilege to do so, is liable to the other for misappropriation ....” Ala. Code § 8-27-3 (emphasis added). Similarly the DTSA defines “misappropriation” as “disclosure or use of a trade secret of another without express or implied consent by a person ....” 18 U.S.C. § 1839(5). Applying either of these definitions to Plaintiffs' wrongful dissemination claim, Plaintiffs must establish that Defendant disclosed a trade secret. Even assuming Defendant supplied a “missing piece” that could allow webinar attendees to derive or ascertain Plaintiffs' trade secret, that is not the same as disclosing that trade secret. The legal point is as tautological as it is true: to be liable for disclosing a trade secret, a defendant must have disclosed that trade secret. Interpreting the ATSA, the Alabama Supreme Court held that a computer programmer was not liable for disclosing the trade secret of his old company's process when he mentioned that process to a coworker at his new job because that coworker was already familiar with the process. Systrends, Inc. v. Group 8760, LLC, 959 So.2d 1052, 1072 (Ala. 2006). In that case, the plaintiff “relie[d] on the cumulative effect of inferences it says are derivable from [an] e-mail and other e-mails to argue that the jury ‘could have reasonably inferred that [the defendant] disclosed and [the defendant's employer] used [the plaintiff's] trade secrets.” Id. Although not completely on point, this case highlights that disclosing a trade secret is not the same thing as disclosing what is allegedly some “missing piece” that might allow the receiver to infer the underlying trade secret.
Because there is undisputed evidence that the webinar notebook did not disclose Plaintiffs' second alleged trade secret, and because there is no legal basis for finding an indirect disclosure occurred here, summary judgment is also due to be entered in favor of Defendant as to this trade secret.
iii. Plaintiffs' Third Purported Trade Secret
Finally, it is undisputed that the January 2021 Webinar did not disclose either explicitly or implicitly Plaintiffs' third alleged trade secret: displaying the data in a customized way. (Docs. # 166 ¶ 33; 178 at 14). The only functions that were displayed and disseminated in the webinar were Wolfram's own built-in functions. (Docs. # 166 ¶ 33; 178 at 14; 164-3 at 17, 21, 23).
While Plaintiffs acknowledge that the functions within individual lines were not customized, they imply that the entirety of the thirteen lines and their functions could be considered a customized collective (rather than individual) function for displaying data. (Doc. # 178 at 14) (“Undisputed for the purposes of summary judgment that Wolfram did not utilize any customized individual functions in the January 21 webinar notebook.”) (emphasis in original). But that assertion misses the mark. Plaintiffs provide no elaboration on this argument (and at most only imply it). Defendant cannot manufacture a dispute over whether there might have been such a disclosure. Defendant is entitled to summary judgment as to this trade secret claim as well.
V. Motion to Bar Expert Testimony
As noted above, Defendant also filed a Motion to Bar the Expert Testimony of W. Anthony Mason (Doc. # 161). This Motion argues that Mason lacks relevant qualifications to offer opinions as to the value of Plaintiffs' alleged trade secrets as well as whether they were generally known or readily ascertainable. (Id. at 2-6). It further argues that Mason's testimony as to memoranda, emails, and social media posts should be excluded because it is not based on a reliable methodology. (Id. at 6-10). The court does not analyze the first two of these arguments because none of the court's above analysis relied on this evidence.
Defendant's third argument is that Mason's testimony about the “copying” of cells from the BloombergTerminal.nb notebook into the webinar notebook is not helpful to the trier of fact because it only analyzes metadata and not the content of the cells. (Id. at 11-14). The court disagrees.
Defendant does not challenge Mason's qualifications to offer this opinion, nor does it challenge Mason's methodology. The court notes that Mason is qualified to offer this opinion, having published “peer-reviewed original research on reducing plagiarism in a computer programming course.” (Doc. # 164-8 at 8). Moreover, Mason uses a reliable methodology to identify copied cells, using Python code to extract CellIDs in various notebooks and listing cells with overlapping CellIDs. (Id. at 32, 37-38).
Defendant argues that Mason's opinion is unhelpful because it could tend to confuse the trier of fact by focusing only on the fact that the lines were copied, and not on whether the content of the lines were trade secrets. (Id.). But this argument cuts no ice at all.
Defendant next argues that this copying evidence is irrelevant because it is undisputed that these lines individually were not trade secrets and it does not make it more or less likely that a trade secret was copied. (Id. at 11-14). Although Defendant is correct that none of these lines individually constitute trade secrets, the fact that four lines of code in the BloombergTerminal.nb notebook had the same CellID as four lines of code in the webinar notebook is notable.
Under Federal Rule of Evidence 401, “Evidence is relevant if: (a) it has any tendency to make a fact more or less probable than it would be without the evidence; and (b) the fact is of consequence in determining the action.” Fed.R.Evid. 401. Applying this definition, evidence that four lines of code were copied from the BloombergTerminal.nb notebook into the webinar notebook is relevant. First, the fact that a Wolfram employee had copied lines from the BloombergTerminal.nb means that the employee had the notebook open at the same time as (or shortly before) they had the webinar notebook open. This fact, in turn, makes it more probable than it would be without the evidence that the employee also copied any trade secrets that were present in the BloombergTerminal.nb notebook into the webinar notebook. This is because the Wolfram employee would have been able to view any trade secrets in the BloombergTerminal.nb either shortly before or while the webinar notebook was open, making it easier for that employee to copy such secrets. Second, if established, the “fact” that a Wolfram employee had copied such trade secrets into the webinar notebook is of consequence in determining this action. This is because it is undisputed that the webinar notebook was distributed, and therefore, the “fact” of trade secrets being copied into the webinar notebook from the BloombergTerminal.nb notebook was considered by the court in evaluating Plaintiffs' trade secret dissemination claim. Therefore, this aspect of Defendant's Motion to Bar Expert Testimony is denied.
VI. Conclusion
For the reasons discussed above, Defendant's Motion for Summary Judgment (Doc. # 163) is due to be granted, and Defendant's Motion to Bar Expert Testimony (Doc. # 161) is due to be denied. An order consistent with this memorandum opinion will be entered contemporaneously.
DONE and ORDERED.