We have access to more information today than ever before. Our experts combine tried-and-true econometric methods with sophisticated data analytics capabilities to enhance the deep insights we provide to clients. Using innovative technologies – including proprietary modeling techniques – we apply modern analytical approaches to review structured and unstructured datasets, optimizing standard, academically rigorous analyses. Our projects are enhanced by integrating data analytics tools into projects across all of Brattle’s practices.
Data Analytics at Brattle
Decoding complexity with technology
We apply modern analytical approaches to projects of all sizes to optimize classic economic, financial, and regulatory analysis.
Growing with Data:
Economics at Scale
Competency Areas

Big Data Analysis

Machine Learning & Generative AI

Data Visualization Platforms

Geospatial Analytics

Document &
Text Analytics (NLP)

Interactive Simulation & Modeling

Social Media and Network Analytics

Asset Tracing – Banking & Blockchain

Data
Engineering
Leveraging state-of-the-art technology such as virtual machines and clustered computing, we excel in handling, processing, and analyzing colossal datasets, up to and including petabytes of data. At Brattle, we ensure the efficient transfer and management of these datasets, which are crucial for the sophisticated analyses required by our clients. Understanding that every project is unique, we tailor our tools and processes to meet specific needs, ensuring both immediate effectiveness and long-term sustainability. We can run thousands of parallel Monte Carlo simulations and other sophisticated analyses required by our clients to empower data-driven decision-making.

Big Data Analysis
Machine learning and generative AI algorithms are not a substitute for human reviews but a powerful supplement, able to dramatically reduce scope and effort in order to prioritize human expertise where it is most needed. At Brattle, we leverage these advanced tools to categorize, de-duplicate, and detect patterns in data, even generating content such as metadata values when necessary. These techniques also bolster our advanced optical character recognition (OCR) platform, enabling accurate information extraction from low-quality or corrupted documents. Additionally, these technologies enhance our flexibility in tasks like web scraping, allowing us to adapt to dynamic online environments and uncover valuable insights with greater efficiency.

Machine Learning & Generative AI
Data visualization tools have revolutionized the way we interact with data, making complex datasets more accessible and understandable. At Brattle, we use these tools to transform complex datasets into compelling stories, enabling users to identify patterns, trends, and outliers swiftly. By combining data visualization with the flexibility of code, we create highly customizable, dynamic charts and graphs tailored to specific needs.

Data Visualization Platforms
Brattle uses geospatial analytics in a wide range of applications, including utilities planning, environmental litigation, telecommunications projects, and M&A proceedings. Mapping information is critical when working on overlap analyses or calculating draw areas, and this data can also supplement hedonic pricing models and analyze the effects of environmental contamination. Combined with our capabilities in data visualization, geospatial analytics can help transform data into compelling narratives.

Geospatial Analytics
Text and document analytics tools – such as PDF scraping, eDiscovery, sentiment analysis, and entity recognition – have transformed the way we analyze and make sense of unstructured data, such as emails, social media posts, and other documents. Commonly relied upon by our White Collar Investigations & Litigation practice, these tools can pull financial figures, email metadata, and other key information to enable clients to make informed decisions based on large document productions. Whether the client is looking for critical documents, constructing a timeline, or identifying themes, our deep experience will ensure that the insights needed are revealed efficiently and accurately.

Document & Text Analytics (NLP)
Along with expert analysis, Brattle provides tools that provide clients with the ability to answer ad hoc questions about data. This may take the form of interactive dashboards capable of rendering millions of records, HTML web pages with embedded custom R and Python visuals, or interactive simulations such as those created with our Grid Scenario Impact Model (gridSIM). Learn more about the unique applications of modeling in Brattle’s Electricity practices here.

Interactive Simulation & Modeling
As a key driver in the rise of big data, social media data presents unique challenges in litigation, including those related to proprietary data formats and business logic. At Brattle, we utilize advanced social media and network analysis techniques to uncover critical insights in complex legal matters. Our expertise enables us to identify patterns, networks, and evidence of conspiratorial activities, providing robust support for litigation and other analytical needs.
For example, analyzing the open forum of messages on platforms like Discord requires a different approach than traditional direct communication channels such as email and phone. Our experience with these technologies has allowed us to develop programs and logic to transform massive communications data into comprehensible network analyses. In one pro bono case, we processed and analyzed over 32 million Discord messages to uncover networks and patterns of interaction, revealing conspiratorial activity and providing critical evidence that played a significant role in the trial’s outcome.

Social Media and Network Analytics
Asset tracing is an important tool in banking and cryptocurrency litigation, where it is used to track the flow of funds and financial or digital assets between parties. Tracing can identify the source of fraudulent transactions, trace hidden assets, and provide evidence of money laundering activities, and it is important in ensuring transparency and accountability – particularly with the anonymity and decentralization of crypto transactions. The flow of currency (both fiat and crypto) can reveal patterns and relationships between parties and provide crucial evidence in legal proceedings. This is why we have developed numerous tracing algorithms through work in our Cryptocurrency & Digital Assets, White Collar Investigations & Litigation, and other practices

Asset Tracing – Banking & Blockchain
Brattle’s Data Engineering team plays a crucial role in our projects, from the initial data transfer to the final report, trial preparation, and beyond. With robust coding skills and extensive experience, our team ensures that each project utilizes the most effective technologies available. In addition to their project-specific contributions, the Data Engineering team is dedicated to researching and implementing cutting-edge technologies to enhance project efficiency and introduce new capabilities. They also create comprehensive documentation and training programs, ensuring that expertise in the latest tools is shared across the firm, thereby reinforcing our commitment to continuous improvement and innovation.

Data Engineering

Selected Engagements
This Competition practice project led by principal Minjae Song, involved an econometric analysis that used a database of transaction records totaling approximately 30 Terabytes. The team took the following steps to conduct the analysis efficiently without compromising its robustness:
- Utilized cloud computing tools to analyze transactions associated with over 150 million accounts.
- Leveraged parallel processing to reduce data processing times by over 50%.
- Implemented random sampling techniques to estimate an econometric model 75 times.
In Re Apple iPhone Antitrust Litigation
This project used Brattle’s proprietary capacity expansion model, GridSIM, to evaluate the role of battery storage in ISO New England’s electricity market through 2050. GridSIM optimizes the energy resource mix given future market and policy conditions, such as decarbonization policies and electrification-driven load growth. The project team simulated the ISO-NE market and evaluated battery storage market potential under a range of future conditions by taking the following steps:
-
Creating an interactive Power BI dashboard to visualize and efficiently interpret key market results from each GridSIM model run, such as capacity buildout, generation mix, and hourly energy prices for each modeled year through 2050
-
Calibrating and benchmarking each model run against recent ISO-NE market prices and other public studies to validate model results
-
Sensitizing key model inputs (e.g., renewable energy incentives) to evaluate battery storage in a range of future conditions
Electricity gridSIM Modeling
Brattle was retained by an insurance broker to develop a tool to predict settlement amounts in securities class action cases; the tool would be integrated into a platform that helped the broker’s potential clients determine which level of Director & Officer (D&O) insurance coverage to purchase. The tool, which predicts settlement amounts based on regression analysis, includes a list of key variables for the regression but also allows the user to include additional variables if desired. After the user has selected the variables to be included, the tool will run a regression and display the result. Based on the regression results and other inputs from the user, the tool then calculates the predicted settlement amount and its probability distribution. Graphics will also be generated to illustrate the results. All of these happen in real time when the user selects the regression variables to be used and other inputs for settlement prediction—results and graphics will be updated on the fly. The tool is hosted on the cloud securely to be accessible to authorized individuals from our client.
Insurance Price Modeling
In a tax dispute, a Brattle team was retained to track the source of a particular part – each identifiable by a unique serial number – installed in the final product in over 12,000 sales transactions. This required the team to extract the serial numbers of the part of interest from sales records produced by the opposing side, and then to use those serial numbers to trace how the opposing side acquired those parts. The produced data was voluminous and unstructured – Brattle received over 100,000 documents in different formats, some of which were hundreds of pages long and with transaction records presented in different layouts.
A major part of our analysis was, for each of the 12,000 sales transactions, to efficiently identify the relevant subset of documents to review from this massive data set. To that end, the project team:
-
performed OCR and extracted full text from the produced documents
-
cleaned up and reduced the size of the full text data so that any keyword search can be performed faster
-
wrote Python scripts to identify documents that contain the keywords or text patterns that the legal team was interested in.
Manipulation of Inventory Receivables
In Re Apple iPhone Antitrust Litigation
Electricity gridSIM Modeling
Manipulation of Inventory Receivables
Insurance Price Modeling
Selected Engagements
Frequently Asked Questions
What is Brattle's experience with big data?
Big data can take many forms, and what used to be considered "big" is now routine. For example, Brattle regularly handles terabytes of documents, and has experience analyzing petabytes of tabular data.
Is it safe to store data in the cloud?
Data stored in our Azure cloud environment meets the strictest level of data security including:
If a client has a specific security standard with they need to meet, they can refer to this list which covers all of Azure’s security standards compliance.
How can I securely transfer data to Brattle?
Data can come in different shapes and sizes, necessitating a range of delivery methods. We've built a diagram to help determine which delivery methods best suit which types of data. CLICK FOR MORE DETAILS
How is Brattle using generative AI?
Brattle applies generative AI across various projects, including research, document review, process automation, and advanced analytical workstreams. We incorporate the latest large language models (LLMs) to enhance efficiency while maintaining high standards of accuracy. Our internal audit processes ensure rigorous oversight, allowing us to validate AI-generated outputs and mitigate potential risks. Refer to our Selected Engagements for some detailed highlights.
SFTP File Transfer
Too big for email < 50 GB
- Brattle can configure an SFTP Account to facilitate two way data transfers
-
This classic approach is simple and reliable
Citrix ShareFile
<1 TB
-
Brattle uses the Citrix ShareFile service to transfer large datasets
-
There are still some limitations such as an individual file size limit of 100 GB
-
This tool expands the functionality of SFTP
Physical Hard Drive
1 TB–16 TB
-
Brattle can recieve devices at the Boston office
-
This is a reliable approach to transfer datasets of any size
Cloud Data Transfer
16 TB+
-
For Data originating in the cloud Brattle has tools to read from AWS, Azure, GCP and more
-
This highly efficient and scalable solution is the best way to transfer the largest datsets
Frequently Asked Questions

Featured Insights






