Websites, web services, bots, integrations and automation. I take the whole task — from framing it to a working version in production — and stay accountable for the result alone, with nothing handed between contractors.
What people usually arrive with
Three routes, one accountable system. Pick the one closest to your task.
OutcomeA working MVP with early users and a clear plan for the next iterations.
OutcomeA repeatable process with connected systems, clear outputs and visible error handling.
OutcomeA diagnosis, prioritised plan and stable working version without a reflexive full rewrite.
Six questions in plain words. The output is a finished brief you can take to any developer. No sign-up, no email, no call.
The tool never quotes a price: without inputs that would be a made-up number. It shows what the price depends on instead.
Commercial projects under NDA: I show context, decisions and outcome — never client data.
The full product path: research, architecture, MVP, first users, first sales and enterprise integrations.
Outcome The product went from hypothesis to first sales and runs as SaaS. Product, architecture, development, infrastructure BackendFrontendPostgreSQLREST API+2Safe evolution of an active enterprise platform: administration, plans, branding and synchronisation without breaking legacy workflows.
Outcome New features shipped into a live multi-client system without downtime. Frontend and backend within the product team AngularAngularJSTypeScriptREST API+4Development and maintenance of web interfaces for an active telecom platform: tables, filters, REST APIs and a mixed legacy stack.
Outcome Legacy areas improved incrementally, without stopping a critical system. Frontend, REST API integration, code review AngularAngularJSExtJSTypeScript+4Every stage ends with a verifiable result rather than a promise of the next stage.
The boundaries are stated upfront — those are the terms.
Not a manager with an invisible subcontracting chain, but the person who sees and owns the whole outcome.
If the scope grows, I stop and explain why rather than silently expanding the invoice.
Intermediate versions and verifiable stages instead of a single big reveal at the end.
Commercial work is never disclosed without approval: cases are anonymised and their boundaries are explicit.
Earlier engineering work includes signal-processing algorithms, microprocessors, calculations and technical documentation. It is not marketed as a mainstream service; it demonstrates an ability to go deeper when a problem requires mathematics and systems thinking.
An exact figure without inputs would be a made-up number. Cost depends on the number of external systems, how much data has to move, the reliability bar and how much interface external users need. An estimate follows a short review of the task, before any work starts, and is fixed together with the scope of the first version.
You do. Source code, repository and access to servers and services are handed to the client rather than kept by the developer. That applies to intermediate versions too: you are never held hostage mid-project.
Yes. Small tasks need an agreed scope and payment order in writing; larger ones get a contract with stages and acceptance. I sign NDAs without objection.
Intermediate versions are shown as the work goes, not at the end, so a mismatch surfaces while it is still cheap to fix. Defects within the agreed scope are fixed at no extra cost. If the task turned out to be different from what we fixed, we discuss it rather than quietly rebuild.
You get the code, the access and a short guide to running and updating it. From there either you take it forward yourself, or we agree support separately. I do not engineer a dependency on me.
A review of the task usually within a day or two. The start of delivery depends on current load, and I am upfront about it: if the next window is three weeks away, you hear that immediately rather than after agreeing terms.
Yes, and for an unfamiliar or risky project that is the right order: ScopeMap, then a short review or diagnosis, and only then a decision to start.
Yes. I first reproduce the issue and inspect the structure. I suggest a full rewrite only when it is demonstrably safer and less expensive.
Writing code, reviewing a large third-party repository, full screen sets and detailed estimates are paid work. The free step demonstrates the reasoning, not a portion of the order.
Start with ScopeMap or describe what happens today and what should happen instead. I will suggest a safe first step.