What we build
Data models that survive product changes, index strategies that keep p99 latency flat, replication topologies for HA, ETL/ELT pipelines into analytics warehouses, and pgvector for RAG.
Where it shows up
Every project. Behind every dashboard and every AI answer is a schema someone had to design carefully.
Example
-- Covering index for the courier "next drop" query
CREATE INDEX CONCURRENTLY idx_drops_active
ON drops (courier_id, status, scheduled_at)
INCLUDE (order_id, address_id)
WHERE status IN ('assigned', 'in_progress');