Engagements

Founder-led builds with a fixed outcome and a real deployment at the end.

I take on a small number of engagements at a time. The work ends with running code, verified infrastructure, and a handoff you can actually operate.

27 Apps live on the same platform spine

700+ Enterprise sites shaped by the same operating discipline

What I build

Three ways to work together.

Each engagement is scoped by deliverable, priced by outcome, and verified in production before it is considered done.

Most common

Production Platform Build

SaaS foundation on Cloudflare Workers and Neon Postgres with auth, CI/CD, observability, and a real custom-domain deployment.

Strict TypeScript API and edge-native runtime decisions
Database, migrations, and row-level security from day one
Runbook, health checks, and rollback-aware delivery

Typical delivery: 3 to 6 weeks

Model product layer

Model Pipeline Integration

Streaming, tool calling, provider failover, and cost controls wired into the product you already have.

Anthropic, Grok, and Groq cascade support
Fail-closed gateway validation and spend discipline
Production-grade SSE and agent loop scaffolding

Typical delivery: 2 to 4 weeks

Media automation

Automated Video System

Template-triggered video generation using LLM scripting, ElevenLabs narration, Remotion rendering, and Cloudflare Stream delivery.

Script, audio, render, and publish pipeline
GitHub Actions driven automation with monitoring
Built for repeatability, not one-off launches

Typical delivery: 3 to 5 weeks

Process

Straight line from first conversation to handoff.

No black-box dev cycle. No vague retainer. No mystery around what “done” means.

Fit check

You tell me what you are building and what is blocked. I tell you whether it is a fit, what I would scope, and where I would not take the work.

Scoped outcome

You get a short scope document with deliverables, timeline, fixed price, and explicit out-of-scope boundaries.

Build in the open

The work happens in a repo you own, with visible pull requests, real deployments, and no dependency theater.

Production verification

Done means the system is live and healthy. If the custom domain cannot answer cleanly, the work is not finished.

Operate without me

You leave with the runbook, secret rotation path, monitoring context, and enough clarity to keep moving after handoff.

Fit

Best for deliberate builders.

Good fit

Founders who want a real backend instead of a future rewrite
Teams shipping model features without infra support around them
Media businesses that need repeatable production instead of manual pipelines
People who value public proof, operating clarity, and lean infrastructure

Not a fit

Projects that want a fast prototype with no production intent
Teams that need a full-time embedded employee
Work that depends on vague scope and endless revisions
Architectures that assume the real cleanup can happen later

Why trust the work

Built by someone who has been responsible for systems people depend on.

Fortune 500 UC infrastructure, disaster recovery, federal VoIP migration, and 25+ years of live operations shape every choice I make. That is why the platform work is fast without being careless.

The Factory platform itself is the proof layer: 27 apps, 130 workflows, public merge history, lean infrastructure cost, and a founder page that shows the receipts.

See the Founder page →

Start with a clear conversation.

Tell me what you are building. I will tell you whether it is a fit, what I would scope it to, and where I would be careful.

Email directly Use the contact form