LLM-powered Office add-in for automated contract review — from prototyping to market launch using OpenAI APIs, embeddings, and vector search with seamless Office integration.

Legal teams spend enormous time manually reviewing contracts for risks and anomalies. Existing tools lacked the semantic understanding needed to reliably identify complex contractual obligations and risks in Japanese and English documents.
Built an Office add-in integrating OpenAI APIs with a custom vector search pipeline. The system embeds contract clauses, retrieves relevant precedents, and uses LLMs to generate risk assessments — all seamlessly within the familiar Office environment.
GPT-powered contract clause analysis identifying risks and obligations using few-shot prompting.
Embedding-based retrieval of relevant contract precedents and similar clauses for context-aware review.
Seamless integration with Microsoft Word via Office.js, providing inline suggestions without context switching.
Led product from prototype to market launch as sole tech lead.
Integrated OpenAI APIs, embeddings, and vector search for semantic contract analysis.
Expanded and led the engineering team post-launch.
Bridged LLM research with product strategy for continuous improvement.