Full-stack engineer
Most of my work is analytics and operations software in Next.js, React, and TypeScript. These days that includes AI features like LLM pipelines and MCP servers. I write code through Claude Code, with my own skills, evals, and verification checks.
Selected work
A few projects that show how I work across analytics depth, operational speed, and trust-heavy customer experiences.
Ecommerce analytics
Multi-tenant analytics for profitability, LTV, and marketing decisions.
Customer-facing analytics, AI-assisted insights, an MCP server for merchant AI clients, and internal tooling across integrations, exports, and complex workflows.
Manufacturing operations
Real-time platform for factory-floor planning, monitoring, and response.
High-density dashboards and workflow surfaces built to stay legible in fast-moving factory environments.
AI math tutoring · side project
Turns one homework problem into a short lesson a child solves on their own.
A parent photographs or types a math exercise. The app recognizes it, generates a worksheet, has a second model review it, and sends the child a link once the parent approves.
Fintech mobile product
Local currency and IOU platform built for community-scale financial trust.
bekiPay enables local currency payments in Luxembourg; Sikoba handles peer IOU tracking with automated clearing — both built on a shared React Native foundation.
How I work
I map the decisions a screen needs to support before thinking about polish, so the interface earns its complexity instead of hiding it.
From data shape to UI states, I follow the whole path so integration issues get solved early and the product feels coherent.
I prefer small, working milestones that create feedback loops quickly instead of long stretches of hidden progress.
Good async notes, explicit trade-offs, and honest status updates are part of the delivery work, not extra admin around it.