Can you tell us about yourself and what inspired you to found DeepJudge?

I’m originally from North Macedonia and moved to Switzerland to pursue my higher studies. I completed my Master’s degree in Computer Science at EPFL and later earned a PhD in AI from ETH Zurich, focusing on deep generative models and semantic understanding of text. My professional experience includes working at Google Assistant and Google Brain. I witnessed firsthand the rapid evolution of AI, particularly in text processing both in research and academia as well as industry.

Recognizing the potential impact of these advancements and the big inflection point, my co-founders and I decided to use the momentum and apply the latest breakthroughs to the legal domain through DeepJudge. We were immediately drawn to the unsolved problem of connecting users to their entire collective knowledge in an easy, intuitive and efficient way.

DeepJudge provides an AI-based knowledge search of all the documents in a company. What were some of the challenges you faced in creating such a product and how did you overcome them?

Building a search engine for internal documents comes with numerous hurdles, particularly in accommodating diverse document types (contracts, memos, briefs and even emails), ensuring scalability, and upholding stringent security standards while dealing with a lot of duplicates, versions and lack of structured information. We found existing horizontal and open-source search engines lacking in performance and speed. To overcome this, we built our proprietary technology and AI models tailored specifically for legal documents, delivering an intent-based, Google-like experience within law firms or corporates. DeepJudge’s solution is scalable, accommodating any volume of documents, and can be deployed on-prem or in secure cloud environments while respecting access rights.

What unique aspects of DeepJudge’s AI technology set it apart from other solutions in the legal tech space?

DeepJudge’s AI technology boasts several unique aspects that distinguish it from other solutions in the legal tech space.

Firstly, DeepJudge stands out for its scalability and security measures, ensuring that users can query their entire accumulated knowledge base using natural language while maintaining data integrity and confidentiality. Additionally, DeepJudge is tailored specifically to the legal vertical, offering a user-friendly experience with seamless integration into internal document sources. Its ability to handle unstructured data is one of our very appreciated features, providing AI-predicted metadata to transform document overload into an easily accessible library of prior work products.

Moreover, DeepJudge serves as a key and foundational component of the tech stack, not only providing end users with relevant information but also supporting other AI models with crucial internal data when generating answers. As part of its offering, DeepJudge includes an Assistant that uses the Search as a human would to access underlying data and perform high-level tasks. In the realm of generative AI, DeepJudge stands out for enabling live connections to internal documents while respecting access rights and citing sources. This ensures the integrity and reliability of generated content, making DeepJudge a powerful tool for legal professionals leveraging generative AI technology.

Who are the clients you work with and what segments do you operate in?

We primarily serve law firms and in-house legal departments. However, we also cater to adjacent industries facing similar issues when looking for the right information within a maze of document sources. Our operations initially began in Switzerland, leveraging our multilingual capabilities to assist customers utilizing languages beyond English in the region. Currently, we are expanding our reach to the UK and USA markets, where we already have built an exceptional commercial team.

Technology and AI are developing very fast. How do you see the future of DeepJudge?

Looking ahead, the future of DeepJudge is incredibly exciting. We firmly believe that the question isn’t whether AI should be utilized by law firms and organizations, but rather how it should be integrated. At DeepJudge, we are convinced that an area where AI will really make a difference is in how we locate, utilize, and repurpose prior work, especially given that knowledge work rarely starts from scratch.

DeepJudge serves as the foundation for organizations to prepare for the forthcoming advancements in generative AI. We understand that if humans struggle to find the right information, AI will face the same challenge. To actually benefit from the power of these technologies and go from “it’s fun” to “it’s truly useful”, one needs to pair them with efficient retrieval, navigating efficiently through the mess to find the most important information, which is exactly what our search does. We see a lot of firms looking at DeepJudge as a true investment piece to set their organizations for success in the era of AI.

You were recognized on the Forbes 30 Under 30 list – what does that mean to you and has it had an impact on DeepJudge?

Being recognized on the Forbes 30 Under 30 list was truly an honor and a validation of our team’s dedication and efforts. It provided significant visibility for DeepJudge, putting us on the radar of investors and prospects alike. Personally, I place a lot of value in the concept of role models, and if my recognition encourages even one woman with or without a tech background to pursue entrepreneurship, I consider it a very meaningful achievement.

How did you hear about LEXR and can you describe your experience of working with our team on commercial contracts?

We were introduced to the LEXR team when we won an award at >>Venture>>. Since then, our collaboration with LEXR has been really close-knit. LEXR has proven to be an invaluable asset to startups like ours—agile, flexible, and swift in their operations, all while maintaining a high standard of quality. We rely on LEXR a lot, particularly for our commercial contracts, and have consistently found their services to be efficient and effective.

What advice would you give to aspiring entrepreneurs aiming to innovate in the tech or legal industry, e.g., with an AI business? Are there any lessons you have learned that you wish you had known at the beginning?

We see consistently that AI is a force multiplier and even small teams can make a huge impact and outpace significantly larger teams. That means that for anyone thinking about it, it’s a really good time to just jump and try to change a problem with a team that is inspired by the same mission.

Regardless of the industry, the importance of selecting the right co-founders and team members cannot be overstated. While products and focuses can evolve, your cofounders remain constant through the ups and downs of entrepreneurship. I’ve been fortunate to have incredible co-founders by my side, and their support has been invaluable.

Keep a tight feedback loop with your customers from the outset. Begin with a minimum viable product (MVP) and iterate based on feedback from the right audience before investing more resources. Understanding your customers’ needs and preferences is crucial for refining your product and ensuring its success in the market.