Expert PostgreSQL Database Engineer (pgvector & High-Concurrency Tuning)

We are building a highly scalable "Enterprise Secure Knowledge-Base Retrieval and Analytics Dashboard." Our system relies on a unified database that must handle standard hierarchical relational data (users, sub-accounts, logs) alongside millions of heavy AI vector embeddings. We need an expert to architect the schema, configure the vector indexes, and implement strict connection pooling to survive massive traffic spikes. The Tech Stack - Database Engine: PostgreSQL - Vector Extension: pgvector - Connection Pooling: PgBouncer - Infrastructure: Google Cloud / Docker Key Responsibilities - Design a scalable database schema that safely partitions heavily updated relational tables (like chat logs) from static, heavy vector embedding tables. - Configure and tune pgvector using HNSW indexing for millisecond-latency cosine similarity searches. - Implement and configure PgBouncer to manage high-volume, concurrent traffic from our serverless Python backend without crashing the database. - Optimize database memory (shared_buffers, work_mem) specifically for vector search workloads. Required Experience - Proven, production-level experience managing heavy PostgreSQL databases. - Direct, hands-on experience building and tuning the pgvector extension (specifically configuring m and ef_construction parameters). - Deep understanding of connection pooling architecture (PgBouncer). - Experience packaging database configurations via Docker for cloud deployment.