Choosing the right database on AWS is one of the most consequential early decisions a startup makes, and switching later is expensive. This agent eliminates the guesswork by mapping your application type, access patterns, and expected scale to the right AWS database from the start. It covers the full portfolio - Aurora, RDS, DynamoDB, ElastiCache/Valkey, OpenSearch, Neptune, MemoryDB, Timestream, DocumentDB, and DSQL - and knows when each one fits based on your specific workload. If you're building AI features, it guides you through vector database selection: pgvector for small workloads, OpenSearch for scale, S3 Vectors for massive corpora, with clear trade-offs at each tier. Once you've picked a database, it doesn't stop there. It advises on high availability and disaster recovery at every level, from free multi-AZ storage replication through multi-region active-active architectures, always with cost trade-offs so you're not over-engineering for your stage. For cost optimization, it walks through Savings Plans, Reserved Instances, I/O-Optimized storage decisions, and scale-to-zero configurations so you only pay for what you use. When you outgrow your current setup, it handles heterogeneous migrations - Oracle to RDS, MSSQL to Aurora PostgreSQL - with DMS patterns, major version upgrade planning, and failure recovery strategies. It also covers day-to-day operational concerns: connection pooling with RDS Proxy, Lambda cold-start patterns, PostgreSQL performance tuning, partitioning strategies, CDC/zero-ETL pipelines to analytics, and multi-tenant schema isolation. It uses the AWS Knowledge MCP server to look up live documentation, so answers reflect current service capabilities rather than stale training data.