Legal IT Latest: ‘AI copilot’ Spellbook raises $10.9m, Travers launches open source ChatGPT alternative, plus all the latest gen AI announcements 

North American ‘AI copilot for lawyers’ Spellbook (formerly Rally) yesterday (25 May) announced $10.9M in funding led by Moxxie Ventures, as well as a formal rebrand from Rally to Spellbook. The round included investment from Thomson Reuters Ventures, Inovia Capital, The LegalTech Fund, Bling Capital, N49P, Concrete Ventures, Good News Ventures, and Venture Newfoundland & Labrador. 

Spellbook, which launched in September 2022, integrates with Microsoft Word and leverages OpenAI’s GPT-4 and other fine-tuned large language models to enable legal practitioners to draft and review contracts “up to four times faster.” 

Spellbook is in early access, and we’re told that it has over 53,000 waitlist sign ups. Beta waitlists are a recurring issue for law firms lately, with the head of one UK top 50 firm recently observing that it has become a source of frustration and competition between firms. 

Spellbook was co-founded by CEO Scott Stevenson, COO Daniel Di Maria, and chief experience officer Matt Mayers. Stevenson and Mayers come from outside the legal sector, while Di Maria was formerly an articling student at John Wenus & Associates in Toronto.  

“It’s clear that AI will fundamentally transform every industry, and the explosive demand we’ve seen for Spellbook made it obvious that our approach has caught on,” said Stevenson. “This round of funding is further proof of our thesis – we’re excited to give lawyers more time to focus on thoughtful work and avoid drudgery, all while reducing errors and increasing accuracy.” 

“At Thomson Reuters Ventures we’re committed to investing in technology to help accelerate innovation in the applications of generative AI,” said Tamara Steffens, managing director at Thomson Reuters Ventures. “Thomson Reuters plans to explore the application of Spellbook’s generative AI technology to our industry leading, authoritative content to deliver customer value,”  

Elsewhere, Travers Smith LLP has formally launched its latest open source project, YCNBot, in collaboration with 273 Ventures, led by former LexPredict founders Mike Bommarito and Dan Katz.  

Created by Travers’ legal technology team, YCNBot (which stands for Your Company Name e.g. TSBot), is designed to integrate with GPT to facilitate generative chat. Built on .NET 7 and React, it is available for anyone to download and deploy through an open source license, enabling, law firms and other organisations to replace the consumer facing ChatGPT with a chat-bot that plugs into Enterprise API’s of Microsoft and OpenAI.   

TSBot has been rolled out across at Travers Smith, with lawyers and business services employees encouraged to learn, experiment, and provide feedback to the legal technology team. It will not be used for any direct client work.  

Oliver Bethell, Travers’ chief technology officer, commented: “Our team of legal technology and AI experts are focussed on research and development to understand how this incredible new technology could impact legal services in the future. By collaborating with Mike and Dan at 273 Ventures, we feel like we’re really at the forefront of trying to understand how computational law will impact our markets and clients. YCNBot has been a fantastic catalyst for conversations with our clients, to understand how they are viewing the opportunities and risks of generative AI.”  

Other gen AI announcements in brief: 

Colorado-headquartered legal case and matter management software system LawBase announced on 23 May that it has added generative AI capabilities within its platform. 

“We are excited to be able to offer this additional layer of AI in our product,” said Phil Homburger, president of LawBase. “Our team has developed this to help our users get even easier access to their data without compromising the security of that data. We have always had strict security practices in place in LawBase – both functional and for the data. But now, when users are looking for information, if they have been assigned rights, they can write the query for exactly what they need and get access to that information from the system, saving valuable time and getting to the important information quickly. Using the new voice-typing feature built into Windows, a user can just talk to LawBase to find the requisite information.” 


London-founded but now New York-headquartered AI-data extraction company Eigen Technologies also on 23 May announced the global availability of its new integration with GPT. 

This option, available to all Eigen customers, allows users to combine Eigen’s no-code AI-powered platform with GPT. The integration means that Eigen customers can now choose to use OpenAI’s GPT model alongside Eigen’s proprietary machine learning (ML) for intelligent automation and document processing purposes.  

Eigen users will be able to choose between using GPT, or any other large language models (LLM), as an alternative to BERT to enhance Eigen’s Instant Answers and extraction capabilities.  

“We know from talking to customers, partners, and the wider market that while there is huge excitement about LLMs, many organisations are unsure of how best to start using them in a real business context. The Eigen GPT integration is the solution to this challenge,” said Dr Lewis Z. Liu, Eigen’s founder & CEO. “We know from our experience working with complex document and data challenges in the financial services and insurance sectors that it takes a full platform to deliver true automation. That requires domain specific knowledge, clear model governance that includes human-in-the-loop where necessary and end-to-end workflow. Eigen provides all of this along with our market-leading accuracy rates, combined with the power of GPT. For many of our customers this combination is proving to be an excellent way of introducing GPT into their automation processes.” 


Intelligent notetaking and document analysis provider LiquidText has announced an integration with ChatGPT and OpenAI tools that will allow users to ask questions and get answers about their project documents, and automatically create reference citations.  

Founded and headquartered in New York, Liquidtext helps users to annotate and connect the information across multiple documents, useful in the likes of depositions and trials.  

The ChatGPT integration allows users to ask questions about any or all project documents and LiquidText uses ChatGPT to analyze the documents and provide an answer. The second feature allows users to type any sentence into their notes, then LiquidText uses OpenAI semantic analysis tools to suggest a citation, and then links to the parts of the documents users were likely taking notes on.  

“By bringing these features to LiquidText, we hope to harness the large language model technology to help users work smarter every day,” says Craig Tashman, founder and CEO of LiquidText. “These new features have potential to help people who read and analyze large document sets make connections they might not have found through their own analysis.” 

These new features are available to LiquidText Subscribers on a pay-as-you-go credit basis, following OpenAI’s fee structure, so users only pay for the AI resources they use.  

This version is being released as a beta edition for LiquidText iPad subscribers only. The Windows version will be added when the beta is complete.