Arnab Chatterjee

Working on Optischema-Slim - a tool that helps you optimize PostgreSQL without needing to be a database expert.

The idea: most teams have slow queries but don't know why. Reading EXPLAIN plans is painful, and guessing at indexes usually makes things worse.

So we built something that:

• Explains query plans in plain English (not cryptic node types)

• Suggests indexes that actually help (simulates them first)

• Rewrites queries when there's a better way

• Scans for common issues (bloat, missing vacuums, lock contention)

I m building OptiSchema Slim: a local-first, open-source AI tool for PostgreSQL performance. I made it because my Postgres tuning workflow kept being reactive and messy, you notice slowness late, stare at EXPLAIN, guess an index, apply it, and only then find out if it helped. I wanted something that stays private (no data egress), but still gives practical guidance.

It includes a Health Doctor scan (bloat, unused indexes, slow/high-impact queries), plus what-if index testing before you touch schema.

I m doing this completely solo (and with basically $0 budget), so I m nervous about launch day.

Question for the folks who ve launched before:
If you could go back to your very first PH launch, what s the one thing you d do differently?

I m building OptiSchema Slim, a local-first AI tool for PostgreSQL performance. It finds slow/high-impact queries, explains plans, and can test what-if index ideas before you apply anything.

For people who tune Postgres in real life: what would you need to see to trust a recommendation?