eDiscovery can be tough, but can it be cruel? If so, Beagle’s goal is to provide its customers with a “cruelty-free” eDiscovery solution!

I recently met with Sergey Demyanov, CEO of Beagle, who gave me a demo of the product and discussed Beagle’s unique approach to providing an eDiscovery solution.

What is Beagle? According to Sergey, Beagle is the “first AI-native end-to-end eDiscovery platform for document reviews and investigations” that “makes the work of legal professionals ‘cruelty-free’ by automating routine tasks and simplifying user workflows.”

Here are some of the notable features, capabilities and considerations associated with Beagle:

“AI-Native” and End-to-End Platform

Beagle is built from the ground up with AI capabilities in mind. As Sergey noted, unlike other platforms, AI is not a “bolt-on” feature but is integrated into the core of the platform: “’AI native’ means that we designed it with an understanding of what AI is capable of and how it can help with different user workflows.”

The platform handles the entire eDiscovery process, from data processing to production. This includes indexing, searching, data analysis, classification, information extraction, aggregation, text generation, tagging, redactions, productions, and exports.

Differentiation from Competitors

Comparison to other AI-startup platforms: I asked Sergey how Beagle differs from other AI eDiscovery start-ups. He explained that “Beagle” is a standalone end-to-end platform, not a plugin, so it can handle all eDiscovery tasks within a single ecosystem.

Automation approach: One of the most notable differentiators is that Beagle uses a natural language task system that supports arbitrary instructions and output structures, while competitors may focus on specific tasks, ‘stories’ or other less flexible formats.

Comparison to the established eDiscovery platforms: Sergey stated that, in addition to its ease of use and powerful document review capabilities, Beagle can also be deployed in a client’s private cloud environment, whereas other vendors primarily focus on SaaS solutions.

“Cruelty-Free” eDiscovery and User Experience

“Cruelty-Free” eDiscovery: The platform aims to make legal work “cruelty-free,” meaning fast and productive with a highly intuitive user interface. The user interface is designed to be familiar to those with experience in eDiscovery, with a similar look and feel to Gmail.

That screenshot is also available here, full-sized.

UI features: Beagle’s UI has features like a document snippet list, document preview, metadata view, coding panel, and related documents, and provides very powerful filtering capabilities

That screenshot is also available here, full-sized.

Flexible “Job” System

Natural language based: As noted above, a key differentiator for Beagle is its flexible task system, which allows users to define tasks using natural language, essentially capturing review protocol and associated data. Users define instructions, data scope, outputs, and assign tasks to individuals or teams. This allows for the flexibility to adapt to any task in eDiscovery. In Beagle such tasks are called jobs.

That screenshot is also available here, full-sized.

Configuring jobs: Jobs can be configured to extract different types of data, including text, dates, integers, and structured data using nested configurations. As Sergey noted: “You may want to extract key data, like the dates and the description of the events from a case complaint. With jobs, you would be able to configure a job to extract such information in a table format. In the same way you can extract all PII data”

That screenshot is also available here, full-sized.

Job flexibility: Users are not limited by specific tasks (e.g. responsiveness tagging) and are able to accomplish any task that can be described in instructions, including document translation, summary generation, or privilege classification.

That screenshot is also available here, full-sized.

Metadata creation: There is no limit to the number of output fields that can be created via these jobs to capture results, and all of them can be used for filtering, configuring new jobs or productions.

AI and Data Extraction

AI-assisted review: Beagle uses AI to generate explanations and highlights within documents that are helpful for data review to provide helpful information to facilitate review. Users can manually create highlights or rely on AI, offering a flexible approach to document review.

That screenshot is also available here, full-sized.

Redactions: You can choose which highlights should be converted to redactions, and the system unifies redaction creation into one framework across manual or automated processes. As Sergey stated: “You can create highlights in different ways, either automatically or manually, then choose which of them should be redacted and how they should be redacted.”

That screenshot is also available here, full-sized.

PII identification: Beagle provides a special solution to identify and aggregate Personally Identifiable Information (PII) of the same people across documents. This PII is pulled from all documents and aggregated into records for each person which can then be exported, making it ideal to facilitate incident response after a cyber incident. The solution is also capable of handling tables (such as Excel workbooks, CSV files, and databases) regardless of their size. Thanks to LLMs that can understand table structures, Beagle achieves 100% accuracy on such data, according to internal testing.

That screenshot is also available here, full-sized.

Infrastructure and Security

Azure cloud-based: Beagle is hosted on Microsoft Azure Cloud. Each case is hosted in a separate private network with isolated databases and storage to minimize the risk of co-mingling data. Because of this separation, Beagle can also be deployed in a client’s private network.

AI processing: AI processing can be handled through OpenAI directly, OpenAI models within Azure (through either GPT-4o or GPT4o-mini, which is one-tenth of the price of doing it in GPT-4o) or Anthropic.

Zero-data retention: Beagle has a zero-data retention agreement with OpenAI and Anthropic for direct usage, meaning data sent to their endpoints is not stored, a key factor for client data security. Sergey stated: “When we talk about security, we have a zero data retention agreement in place with both OpenAI and Anthropic, meaning that unlike their standard 30 days retention policy, the data that we send to their endpoints is not stored anywhere and is not available for human review.”

Pricing and Hosting

Low hosting fees: Beagle aims to offer a more affordable solution than competitors by adopting a usage-based pricing model for AI – based on the number of processed words – along with very low hosting fees. Sergey stated: “Our business model is based on applying a markup to AI services, but we don’t intend to make money on hosting. Our hosting fees are structured into several tiers based on data volume, designed to simply cover our expenses for Azure infrastructure. Beagle imposes no hosting fees at all when deployed in the client’s infrastructure.

Pricing for AI: AI usage pricing is linked to the text volume of the input data, offering more granular and cost-effective options for short documents versus competitors who often charge per document. An additional fee may be charged if a textual response is requested. The price per unit goes down as the total expense go up. As noted above, Beagle also offers access to the more affordable OpenAI GPT 4o-mini which is 10x cheaper than the default GPT-4o and more powerful GPT-o1-mini, as well as Anthropic models.

Free credit: Beagle also offers a $1000 credit to be used for its AI services to allow people to evaluate them on their data before applying them more broadly.

Company Goals

Short-term revenue goal: Beagle’s immediate goal is to reach 1.5 million Annual Recurring Revenue (ARR).

Fostering use of the platform: The company is focused on exponential growth in case usage and aims to demonstrate the value of its AI services as quality control tools that can gradually replace first-pass review. Additionally, they can be used for other tasks that accelerate the document review process, such as junk removal, data normalization, entity extraction, and more.

To Find Out More

To learn more about Beagle and/or book your own demo, visit the Beagle website or contact sales@ediscoverybeagle.com.

Conclusion

Beagle presents a unique approach to eDiscovery by providing a flexible, end-to-end AI-native platform with a focus on user experience, and most notably, by disrupting the traditional hosting model, positioning themselves for growth and industry disruption. Their flexible, natural language task system, usage-based pricing and “on-premise” deployment options are key differentiators for their product, designed to make eDiscovery “cruelty-free”. They also have the cutest logo in eDiscovery! 😊

So, what do you think? Have you checked out Beagle yet? Please share any comments you might have or if you’d like to know more about a particular topic.

Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.