Reinholds.

Reinholds Bunde

Engineering @ Home

London, UK

I build customer-facing systems for companies that need reliability, speed, and pragmatic engineering.

Currently, I lead engineering at Home, splitting my time between product, JVM backends, distributed systems, and platform reliability.

Previously, I scaled payments infrastructure at Zopa and , built telemetry pipelines at Forter, and led engineering growth at KatKin.

Kotlin Java http4k React Full-Stack AI Agents Observability TypeScript CI/CD

Current Focus & Interests

I am currently optimizing for operational simplicity, highly responsive feedback loops, and workflows that reduce accidental complexity in small engineering teams.

Recent rabbit holes: Practical LLM integration to automate repetitive toil (rather than adding maintenance burden), lightweight state machines on SQL databases, and building deterministic CI/CD pipelines.

What I value: Clear ownership, systems designed to survive real users, and small, high-agency teams shipping code quickly.

Selected Writing & Thoughts

Why I Prefer http4k

It models HTTP as a pure function: HttpHandler = (Request) -> Response. No magic annotations, instant boot, and trivial unit testing.

AI in Real Pipelines

Excellent for eliminating boilerplate, but risky for system layout. Use AI to remove mechanical toil, not engineering judgment, and back it with deterministic tests.

Premature Platform Engineering

Until coordination is a bottleneck, internal platforms are a costly distraction. Prefer simple deployments, boring tech, and maximum product focus.

Engineering Principles & Mindset

Simplicity Over Cleverness

I believe the best system design is the one you can fully hold in your head. Avoid premature abstraction, keep telemetry robust from day one, and favor boring, well-understood technologies that ship fast and stay up.

Product-led Engineering

Great software isn't just about elegant code; it's about solving real human problems. I work closely with product and design, treating engineering decisions as product trade-offs where speed-to-feedback is a primary metric.

Systems Should Be Debuggable

A system you cannot observe is a liability in production. I prioritize clear telemetry, predictable logging, and structured tracing from day one so that debugging is science, not luck.

Prefer Boring Technology

Keep your innovation tokens focused on your unique business problems, not your database or server framework. Choose proven, mature, and boring stacks to keep the operational burden low.