Loading...
Trainee
Trainee at a US law firm in London and a self-taught coder. I believe law is at a key inflection point. Because so much of the work is text-heavy and process-driven, it is uniquely well placed for big gains in speed, cost, and quality over the next few years. I would love to help contribute to that shift. (Recognized on Latham's 2025 AI Honors Roll and winner of the Latham Corporate Harvey AI Competition.)
"The Closing," a gamified simulation engine that transforms Share Purchase Agreements into interactive strategy maps. It uses AI to convert complex clauses (like Indemnities or Earn-outs) into visual game mechanics and simulates "future scenarios" to test if a user’s drafting survives commercial disasters. UI is based on the Matrix (my favourite game growing up).
"The Moot Court," a multi-agent NDA simulator for trainees. It parses NDAs and deploys three distinct AI personas (Opposing Counsel, Client Counsel, and Senior Partner) to debate clauses in real-time. Instead of just auto-correcting text, it teaches trainees the strategy each clause. Built using Python, Flask, and Google Gemini API.
Many “legal AI” tools are not just thin layers on top of frontier models, but actually route users to cheaper, older models.
Contrasts should start with a plain-English summary up front, spelling out the commercial levers and the failure modes, like a GitHub README. The legalese can then sit underneath as the code / execution layer.
Big Law firms that embed agent teams into real workflows, like diligence and signings, will reshape junior lawyering and pull ahead of competitors. Doing that means building serious in-house legal tech capability, with far more technologists / engineers per lawyer than today.