Services
I design and build custom software for technical teams and technical founders: applications, integrations, and technical direction the client owns.
Custom software design and development
I design and build web applications, internal tools, and integrations starting from the operational problem, not the technology. I deliver working software in production, with client-owned code, tests, and documentation.
Example: OpenFatture, e-invoicing without vendor lock-in →Technical coordination and fractional CTO
Part-time technical leadership for companies working with external vendors or an internal team: stack decisions, quote and code review, vendor management, and roadmap. I have led technical teams as an interim CTO and shipped production systems end-to-end.
Background and experience →AI integration and process automation
I integrate AI into business processes where it brings verifiable value: automations with human oversight, agents, retrieval over internal data. Systems that stay inspectable, with audit trails and cost under control.
How I build reliable AI systems →Production readiness audit (1–2 weeks)
Fixed-scope review of an existing LLM workflow: state and recovery paths, eval coverage, cost ceilings, and what to change first. You get a written assessment and a prioritized backlog in your repo — no retainer required to start. (Price on request by email.)
Production AI systems I have built →How I work
Each cycle produces something you can use. The written proposal names scope, timeline, and cost before development starts.
- 01
Analysis
I start from the operational problem: processes, constraints, existing systems.
- 02
Proposal
I put the proposed solution, scope, timeline, and costs in writing.
- 03
Iterative development
Short delivery cycles with working demos: you judge the software by using it.
- 04
Delivery and handover
Production rollout, training, and agreed maintenance.
Frequently asked questions
Project delivery
- Is this a good fit for my project?
- Good fit: technical teams or technical founders shipping an LLM-backed system with a working prototype, where the hard part is production (durable state, deterministic recovery, reproducible evals, or cost under a real budget), and you want client-owned code, tests, and documentation. Not a fit: "add ChatGPT to the website" demos or marketing theater without an operational problem, pure strategy decks with no build or handover, or teams looking for a large agency or a sales-led process.
- Who owns the code?
- The client. I hand over the repository, documentation, and access — no dependency on me or on anyone else.
- How are timeline and costs estimated?
- After the initial analysis I prepare a written proposal with scope, milestones, and costs. Work proceeds in verifiable iterations, not as a black box.
- What happens after delivery?
- I agree on maintenance and support based on what is actually needed: from monitoring only to ongoing improvements.
- Where do we start?
- With an email describing the problem: context, systems involved, and expected timeline are enough for a first assessment.
Fit, systems and proof
- Who do you build software for?
- Technical teams and technical founders shipping LLM-backed systems that must survive contact with production. Usually the prototype already works and nobody owns the job of making it hold up against real traffic, real data, and a real budget.
- Can you work with an existing team and stack?
- Yes. I plug into the stack and constraints you already have. Typical work: hardening an LLM workflow with durable state and deterministic recovery, putting evals around a system before it ships, or bounding its cost — and I leave your team documentation and tests it owns.
- What is it like to work with a single external technical contact?
- You work with one person from requirements to production: I make the stack and architecture choices, write and review the code, and keep durable state, typed tool contracts, audit trails, evals, and recovery paths inspectable. No vendor to coordinate and no team to staff on your side.
- When do you add AI to a system, and when is it worth it?
- Only where it brings verifiable value, with human oversight and costs kept under control. I add AI to a workflow when the output can be checked and the cost is predictable, not as a default. What you get is capability you can audit, not a black box.
- What proof do you have that your systems work?
- Some of my work is open source and verifiable. The emotional-memory project has a Zenodo DOI (https://doi.org/10.5281/zenodo.19972258) with reproducible benchmarks; systems like orka and OpenFatture are public on GitHub. I show qualitative outcomes and the artifacts behind them.
- How do I start a project with you?
- Write me a few lines about the operational problem and the systems involved — that is enough for a first assessment. I reply by email; there is no form or sales call to book.