About
I started as a developer and never stopped being one. Over 25 years I went from PHP back ends and payment gateways to sensor networks over MQTT, and then to designing the cloud behind a smart-home and professional-security platform: 900+ microservicesarchitected at Ezlocareer notes, 2026-10-02 handling 10M+ events a dayarchitected at Ezlocareer notes, 2026-10-02.
I have led teams and delivery for over 20 years, often under a developer or architect title: interviewing 50+ candidates, running full Agile delivery, reviews and onboarding, mentoring 10–15 engineers, and leading a multi-team group of about 30. I have also founded several companies since 2002, sold some, and won and delivered EU-funded projects.
Today I lead product: I turn requests from partners, dealers and executives into PRDs with interactive simulations, and into specifications engineering can build from. I like knowing how things work underneath: protocols, queues, firmware, the bytes on the wire.
On the side I build the AI tooling I wish I'd had: agent harnesses, an MCP gateway, local models and knowledge systems that let AI do real work while a human approves what matters.
ExperienceAll 13 →
- 2026 — Present
Product Manager · Ezlo
Led product development across executives, partners and engineering teams while driving the full product lifecycle and building the internal AI framework used by the organization.
- Dec 2018 — 2025
Software Architect · Ezlo
Technical architecture and cross-team coordination for the company's next-generation IoT platform, working across cloud, firmware, mobile and web while introducing the new platform alongside the existing MiOS / Vera ecosystem.
- 2018 — Present
Software Architect & AI Integration Consultant · Independent contractor
Independent architecture and engineering work combining AI integration, distributed infrastructure and custom business applications.
ShowcaseAll 11 →
longbrowse: a browser that remembers, for AI agents
An AI agent that browses by itself is good at it and expensive: every click is a model call, every morning it relearns the same sites. longbrowse is one container with a real Chromium, logins that stay, MCP and REST for the agent, a live view for people, and hot-reloaded site plugins. The agent learns a chore once, moves it into a plugin, and from then on makes one call. Watch the 4-minute intro on YouTube.
Node.js · Playwright · Chromium · MCP SDK · noVNC · SQLite · Dockerkv-chunk-fs: deduplicating KV-cache snapshots for local LLMs
A local model with a long system prompt and tools saves its whole KV cache to disk after every turn: hundreds of megabytes, rewritten each time, mostly unchanged. kv-chunk-fs is a FUSE filesystem in Go that stores those snapshots as content-defined chunks, so a turn writes only what changed, and restores them at close to plain-file speed. It runs live under every llama.cpp model on my GPU server.
Go · go-fuse · FUSE · desync · llama.cpp · Docker · systemd · LinuxApps that ask for results, not endpoints: a runtime capability graph
Decoupling apps, mobile apps especially, from the APIs behind them. The cloud publishes a machine-readable map of what it can do; the app asks for a result, such as a user's avatar; a small planner inside the app finds the best way to get it, falls back when a step fails, and learns from every client which paths actually work. AI is optional: the deterministic system works without it.
Capability contracts · Graph planning · Telemetry
