Technologist based in Copenhagen, working at the intersection of embedded AI, data, and software engineering — and the people who have to live with the systems we build.
Most of what I ship falls into three buckets: multi-agent AI systems on Azure, ML workloads pushed down to FPGA hardware, and developer-experience tooling that helps teams adopt AI assistants without generating instant legacy code.
| Project | What it is | Stack |
|---|---|---|
| call-out-agent | SOP-constrained voice agent that places outbound alarm calls, follows procedure exactly, and escalates only when needed | Python · FastAPI · Azure Communication Services · GPT-4o |
| SE-agent | Solution-engineer agent with a pluggable tool system, built spec-first with GitHub Spec Kit | Python · MSAL · Dynamics CRM API |
| Financial-Statement-Review-Multi-Agent-Solution | Four specialised agents automating credit and financial statement review, generated from a BPMN diagram | Azure AI Foundry · Container Apps · Teams |
| trending-SoMe-workflow | Collects social signals, finds the emerging outliers, and emails a creative replication brief | Python · n8n · PostgreSQL |
| ML_SME_FPGA | Feedforward neural network primitives for FPGA via Synchronous Message Exchange — my master's thesis work | C# · SME · VHDL |
- Hands-on-workshop-GH-Copilot — a twelve-lab workshop taking teams from everyday Copilot use to agent workflows and quality guardrails
- github-copilot-use-cases — a reference catalogue of Copilot patterns, from code explanation to MCP and custom agents
Python TypeScript C# PyTorch FastAPI React Azure AI Foundry Bicep Docker Kubernetes
Copenhagen 🇩🇰 · GitHub