Brief
You send a few lines about the project, goals and timeline.
Services
Ways we can work together
From a one-off consultation to ongoing product engineering β pick what fits your stage. Clear scope, weekly demos, no surprises.
π
Complex Web Applications
Production web apps when templates stop working. Dashboards, SaaS products, internal tools, real-time systems. There's a point where no-code stops being enough. Your product needs real-time updates, complex state, custom flows, or performance the templates can't give you. That's when you need code β but code built to survive past its launch week. I build web applications where the frontend actually has to work: interfaces used every day by real people, handling real data, at scale that matters.
- SaaS Products β multi-user apps with auth, roles, complex state that stays consistent across tabs and sessions.
- Internal Tools & Admin Panels β CRMs, moderation panels, ops dashboards. Reliability first, performance second, design third.
- Real-time Systems β chats, live dashboards, collaborative interfaces, video/voice. WebSocket, SSE, WebRTC β I've shipped all three in production.
- Data-heavy Interfaces β tables with 10K+ rows that don't freeze, live-updating charts, filters that don't break the URL.
βοΈ
Vibecode Cleanup & Hardening
You built an app with AI. It works on your machine, mostly. But users hit strange errors, data disappears, edge cases break the flow, and deploying updates feels like defusing a bomb. I audit vibecode projects and fix the invisible things: type safety, error boundaries, race conditions, security holes, broken state management. Your app keeps its ideas β it just stops surprising you.
- Audit report β what's fragile, what's dangerous, what's fine
- Refactoring plan sorted by risk vs effort
- Executed fixes with tests covering critical flows
- Handoff docs so future AI sessions don't break what I fixed
π‘
AI Features Integration
Add real AI to your existing product β beyond a ChatGPT button. Every product owner wants "AI features." Most bolt on a chat widget and call it done. I build AI that actually does work in your product: pipelines with retries, structured outputs, tool-calling agents, MCP servers, RAG over your data. TypeScript end-to-end. Zod-validated everywhere. Observable, cost-controlled, and testable β because "the AI is down" is not an acceptable answer to customers.
- LLM pipelines with structured output (Zod schemas, retries, fallbacks)
- Autonomous agents with your tools and APIs
- MCP servers exposing your product to Claude Desktop / Cursor
- RAG over your documentation, catalog, or knowledge base
π―
Prototypes That Work
From idea to clickable, working prototype in 2β3 weeks. You have an idea. You need to show it to investors, test with users, or convince your team. You don't need six months of development β you need something real people can touch by next month. I build functional prototypes with real data, real interactions, and real AI where it matters. Not a Figma clickthrough β a deployed app with a URL you can send to anyone.
- Deployed prototype (Vercel / Railway)
- Design + code iterated together
- Enough polish to demo, honest about what's stub vs real
- Handoff-ready if you decide to keep building
βοΈ
Codebase Maintenance & Continuity
Long-term partner for your TypeScript codebase. Your developer left. Your product is running but nobody dares touch it. You need someone reliable who can add features, fix bugs, and keep the codebase alive without breaking what already works. I take ownership of existing TypeScript/JavaScript codebases: learn them properly, document what matters, and become the person who ships new features without introducing new problems.
- Onboarding week β I read the codebase, document what's undocumented, flag risks
- Ongoing work β new features, refactors, bug fixes, dependency updates
- Weekly async updates so you know what's happening
- Available in your timezone (Europe / late US East)
Not sure which fits?
Send a few lines about your project and I'll suggest the right format β and a price β within 24 hours.
How it works
A lightweight, async-friendly process β from first message to shipped work.
Intro call
A short call to align on scope, fit and the right format.
Proposal
You get a clear plan with milestones, price and timeline.
Kick-off
We start β async-friendly, with weekly demos and updates.
Frequently asked
How do we get started?
Send me a short brief via the contact form or email. I'll reply within 24 hours, and if it looks like a fit we'll set up a quick intro call.
How are projects priced?
Consultations are fixed per session or hour. Product engineering is monthly, scoped to the work.
What if I'm not sure which service I need?
That's fine β just describe your situation. I'll suggest the right format and a price.
Do you work with teams or only solo founders?
Both. I work directly with founders and product teams. For teams I can also run a code review or a hands-on workshop, with notes and a recording for everyone.
What's your availability and timezone?
I'm based in CET and work async-friendly across timezones. New project slots open monthly β the badge at the top shows the next available start.