What we do
We work in small, focused engagements. We will not outstaff you, and we do not pretend to be bigger than we are.
SaaS product engineering
End-to-end SaaS — system design, application work, observability, and deployment. Particular comfort with the operational realities most decks skip: SSO/MFA/passwordless auth, performance and query optimization, the observability and incident-response work that decides whether a SaaS company sleeps at night.
- Hardening a vibe-coded prototype into a system that holds up in production.
- Designing and shipping authn/authz — SSO, MFA, passwordless — without breaking the existing login flow.
- Tracking down the query that quietly went O(n²) at scale and fixing it under load.
- Standing up the observability and on-call runbooks a growing product needs before its first 3am page.
Platform & automation
Infrastructure as code, CI/CD pipelines, deploy automation, and the developer-experience work that quietly compounds. Small investments that pay back across every deploy, every PR, and every on-call shift after. Useful whether you are starting from scratch or feeling the platform slow the product team down.
- Inheriting a CI pipeline that is slow enough to break flow — and going after the real bottleneck instead of paying for more runners.
- Pulling infrastructure that grew up in a cloud console back into version control, without halting the team that depends on it.
- Replacing a brittle release process — manual checklists, tribal knowledge, "ask the one person who knows" — with something a new hire can run on day one.
- Writing the deploy automation a small team has been meaning to build for a year, and never had the bandwidth to.
Applied AI
Production AI inside real products, not demoware. RAG over your data, agents wired into your APIs, in-app natural-language tools that produce real work product, and honest advice on where AI is — and is not — worth the cost. Across local models (Ollama), AWS Bedrock, and the major hosted providers.
- A RAG assistant over your codebase and internal docs, with conversation history and evaluations.
- In-app natural-language tooling that lets your customers produce real output, not just a chat sidebar.
- Replacing a brittle prompt chain with an evaluated, observable pipeline.
- A short advisory engagement: where in your product, support, or developer workflow does AI actually pay for itself?
Technical leadership & code health
Fractional CTO and tech-lead engagements: architecture review, hiring loops, scoping work, and getting an early team out of the weeds. Also the unglamorous wins — taking a legacy codebase and bringing it closer to what the team would have built if they had known then what they know now. Discovery engagements (two-to-four weeks) live here too.
- Quarterly architecture reviews for a small team without an in-house senior.
- A modernization pass that brings a long-running codebase closer to what the team would build if they were starting today.
- A working prototype of a new feature, plus an honest report on whether to ship it.
- Standing up an interview loop that does not waste senior time.
// also: need a marketing site or a self-updatable site? We do those too — usually as a small fixed-scope piece, with hosting we run for you or a hand-off to your own.
Engagement model
Async-first. Most engagements run between two and twelve weeks. We work in Eastern time and we write more than we meet. Project work is fixed-scope; ongoing retainers are billed monthly with a one-month notice on either side.