● PARTNERS

Built on AWS. Designed for scale.

Trubase DB is Postgres rebuilt from scratch in pure Rust on Tokio. S3 is our ground-truth storage layer. EC2 runs our fleet of identical engine nodes. This page explains what we're building and what we need from our cloud partners.

THE FOUNDER

Olamilekan

Founder, backend engineer — building Trubase DB in public.

Olamilekan is rebuilding Postgres from scratch in pure Rust on Tokio, with the backend lifecycle — mint, branch, suspend, wake, dispose — native to the engine, and proving compatibility continuously against PostgreSQL's own 200,000-test suite on a public scoreboard.

01 — WHAT WE'RE BUILDING

The instant serverless backend

Trubase DB is Postgres rebuilt from scratch in pure Rust on Tokio, with the backend lifecycle built directly into the engine. Each backend is a lightweight task instead of a heavy OS process (~10–30MB). A single engine node can multiplex millions of backend tasks.

AWS infrastructure is not an afterthought — it is core to our architecture. S3 IS our primary database storage layer. EC2 runs our stateless Tokio compute fleet.

02 — HOW WE USE AWS

AWS is the foundation of Trubase

Amazon S3

Ground-truth storage layer. Every backend's copy-on-write layers, archived WAL, and timeline metadata live on S3 — this IS the database's persistent tier.

Amazon EC2

The fleet: N identical instances of the Trubase DB engine binary. Nodes hold only caches — no permanent state — and multiplex millions of backend tasks via Tokio.

EC2 Auto Scaling

The fleet scales by awake-backend count and query load. Any node can serve any backend by reading its timeline from S3.

Amazon CloudFront

CDN and edge routing for global backend access. Low-latency connection routing to the nearest compute node.

Amazon Route 53

DNS management for *.trubase.dev wildcard routing. Every backend gets a unique subdomain.

AWS IAM

Fine-grained access control for S3 bucket policies, cross-account access, and tenant isolation at the S3 prefix level.

Amazon CloudWatch

Observability, metrics, and alerting for compute fleet health, S3 latency, backend lifecycle events, and Tokio runtime statistics.

AWS VPC

Network isolation for the compute fleet. VPC endpoints for S3 access to eliminate data transfer costs.

03 — WHAT WE NEED

Partnership request

→AWS Activate Credits — Compute and storage credits to scale our pre-launch infrastructure. S3 and EC2 are our largest cost centers.
→S3 Storage Credits — S3 is our primary database storage layer. Every page in every backend lives on S3.
→Technical Architecture Review — AWS Solutions Architect engagement to validate our S3-native storage architecture at scale.
→AWS Marketplace Listing — Future distribution channel for enterprise customers with consolidated billing.
04 — AI TOOLING

Anthropic Claude access via AWS

We use Anthropic Claude (via AWS Bedrock or AWS Marketplace billing) as a core AI engineering assistant for the engine build. We need platform billing access through our AWS account for Claude API usage.

Claude is critical to our engineering velocity — the build method is corpus-driven AI-assisted engineering, held honest by PostgreSQL's own 200,000-test suite as the executable spec.

Let's partner.

Trubase is built on AWS from the ground up. S3 is our storage layer. EC2 is our compute layer. We're building the future of serverless backends on your infrastructure.

olamilekan@trubase.dev