Building this myself — curious on your take on TAM expansion. Self-hosted streaming solves the Kafka problem for teams, but I wonder:

1. Market size: How many teams actually need self-hosted vs. managed? (Feels like the TAM is smaller than serverless alternatives) 2. Positioning: Are you targeting "we can't afford Kafka" or "we can't trust cloud"? Different buyer psychology. 3. Next moves: Thinking about protocol support beyond Kafka? Or doubling down on that niche? Ask because I'm running validation with 4 frontier models (GPT-5, Sonnet, Opus, Gemini Pro) on infrastructure plays—helps founders pressure-test positioning before they build. If you want a second opinion from actual ML systems instead of just your co-founder's bias, happy to help. No strings attached.

TypeStream looks solid. Good luck with the launch.