Back home
Writing
Notes from shipping GenAI
Technical write-ups on multi-agent systems, RAG in production, and MLOps — drawn from projects I've actually shipped.
What Real-Time Conversational Intelligence Actually Requires to Ship
Real-time conversational intelligence is more than a chatbot with a fast response time. Lessons from building EmotiSphere, Aria, and SlackMind on routing, latency, and memory that make conversations feel alive.
Conversational AIGenerative AI EngineeringReal-Time SystemsSoundHound Houndify Explained: What It Is and Why It Matters for Voice AI
A breakdown of SoundHound's Houndify platform — Speech-to-Meaning, Deep Meaning Understanding, and Collective AI — and how its approach to conversational intelligence compares to custom multi-agent systems.
Voice AIConversational AIIndustry AnalysisWhat I Learned Shipping Multi-Agent AI Systems to Production
Lessons from building LangGraph agent routing, production RAG pipelines, and long-term memory systems across four real projects — Aria, RetailForecastAI, TrustMed-AI, and SlackMind.
Generative AI EngineeringMulti-Agent SystemsRAGMLOps