GenAI Product & Platform Rebuild
Rebuilt an AI meeting assistant from the ground up, integrating modern GenAI capabilities and scaling the platform for enterprise customers.
Technologies Used
Key Features
Project Overview
Anchor AI’s meeting intelligence platform needed a complete overhaul to take advantage of new GenAI capabilities. The existing system was built before the GPT-4 era and couldn’t leverage modern language models effectively. The rebuild had to happen while maintaining service for existing customers.
The Challenge
The original architecture was designed for older NLP approaches and couldn’t efficiently integrate with modern LLMs. Meeting transcription quality was inconsistent, and the system couldn’t scale to enterprise customer demands. Everything needed to be rebuilt - but existing customers couldn’t be disrupted.
What I Did
As founding engineer, led the full-stack rebuild:
- AI Architecture: Designed RAG-based pipeline for meeting intelligence, combining transcription with contextual understanding
- Platform Rebuild: Rewrote core platform with modern stack optimized for AI workloads
- Enterprise Features: Implemented security, compliance, and scalability features required by enterprise customers
- Migration Strategy: Developed parallel-run approach to migrate customers to new platform without service interruption
Outcome
The rebuilt platform delivered 10x faster processing with significantly higher accuracy. The new RAG pipeline enabled intelligent features that weren’t possible before - action item extraction, discussion summarization, and cross-meeting insights. Enterprise customer acquisition accelerated as the platform met their security and scale requirements.