In the legal tech industry, many startups are automating manual processes for lawyers, with some focusing on solutions for intellectual property (IP) matters. One such company is Solve Intelligence, which has developed AI software tailored to assist patent attorneys in drafting patents for IP analysis and generation.
Based in Delaware, Solve Intelligence recently secured $3 million in funding from investors such as Y Combinator, Amino Capital, General Advance, SAV, Translink Capital, and Nomad Capital.
The company’s chief research officer, Sanj Ahilan, and CEO, Chris Parsonson, both with PhDs in artificial intelligence from University College London, recognized the challenges of obtaining patents while working at technology companies like Huawei and Dyson. Chris was also informed by his patent attorney partner about the limited availability of software for patent attorneys beyond basic tools like Microsoft Word.
Solve Intelligence was established in June 2023 by Ahilan, Parsonson, and Angus Parsonson, the company’s chief technology officer, with the product being launched in July of the same year.
Chris Parsonson highlighted the unique nature of patents, stating that they require a blend of technical and legal expertise unlike other legal areas. He emphasized that the algorithms and AI systems needed for patent-related tasks differ significantly from those required in other legal tech domains.
Solve Intelligence’s product is an in-browser document editor, powered by AI, designed to assist attorneys in intellectual property generation. The AI solution can identify novelty and non-obviousness, providing support in the R&D process. According to Parsonson, each approved patent defines the uniqueness and non-obvious aspects of a new technology compared to others in the public domain.
The company’s AI is capable of identifying novel and non-obvious steps, ranking them based on commercial viability, and initiating the drafting process with the IP firm collaborating with the tech company. Moreover, the AI product aids in patent portfolio infringement litigation following the drafting of the patent.
Solve Intelligence has attracted over 25 IP firms across the U.S., Europe, Asia, and South America, with many reporting efficiency enhancements of 60-90%. Despite being in its early stages, the company has already generated recurring revenue ranging from $100,000 to $1 million through the sale of its product on a per-seat subscription basis.
The projected growth of the patent analytics market to $5.18 billion in 2032, from $1.3 billion in 2022, underscores the increasing significance of IP-related technologies.
Confidentiality concerns around generative AI tools, like ChatGPT, have led some potential user attorneys to refrain from using Solve Intelligence’s AI product. The safeguarding of undisclosed inventions, which are highly confidential until they are patented, remains a primary issue.
Parsonson expressed confidence that the concerns surrounding large language model training and inference will diminish over time as people develop a better understanding of AI. He emphasized the company’s use of state-of-the-art security measures, including encryption during transit and storage on enterprise AWS servers.
Solve confirmed that it does not train its AI on clients’ data but utilizes proprietary models, algorithms, and retrieval augmented generation (RAG) approaches specialized for patent drafting and IP analysis. The company is also actively obtaining patents for these methods and collaborating with IP firm clients to use the AI product in drafting patents.
The company’s leadership anticipates widespread daily use of AI by legal professionals within the next five years, highlighting the need for continuously improving products. They emphasized the importance of a skilled technical team capable of delivering features desired by users promptly.
While some companies, such as PatSnap, Cipher, and IPRally, focus on prior art search and technology classification, others like Harvey AI, Robin AI, and Casetext, prioritize legal contract drafting and the incorporation of relevant case law, without a focus on patents. Solve Intelligence distinguishes itself by concentrating on generative tasks related to patent drafting and office action responses.
Parsonson emphasized that their product will extend beyond patent drafting to assist in various stages of the intellectual property lifecycle.
Solve intends to utilize the recent funding to expand its team, address increased demand, and advance R&D for upcoming patent drafting features. These features, slated for release over the next few months, encompass AI algorithms and processes for technical drawing generation and analysis, patent quality enhancement, AI customization to match attorney style, and more.