Technical deep-dives, product decisions, and honest guides — written by the team building it.
A transparent technical walkthrough of how we implemented Supabase RLS across our entire schema using a single Alembic migration — and what we learned about FORCE ROW LEVEL SECURITY vs. ENABLE ROW LEVEL SECURITY.
Our QA audit flagged four security issues. Here's exactly what they were, how we reproduced them, and the specific code changes we made to fix each one.
A honest breakdown of our weekly AI digest feature: what model we use, what it actually costs per run, what you get, and where it falls short compared to reading papers yourself.
We walked through validating a real research question — 'Physics-informed neural networks for turbulence modeling' — using EXPLORE mode. Here's every step, every gap identified, and the actual output.
A real session using DEVELOP's 5-stage pipeline: Gap Analysis → Hypothesis → Literature Review → Draft → Review. We show the actual outputs at each stage, including one false start.
Our architecture decision: async FastAPI backend with Supabase PostgreSQL (via asyncpg + SQLAlchemy), and Next.js frontend. What tradeoffs we made and whether we'd make them again.
Full article content coming soon.
Want to be notified? Follow us on @stemind_io