Selected work
Proof, not promises.
Two live engagements, anonymised. Each is the same thesis under real production load: domain-driven architecture that gets cheaper to extend, not more expensive. Outcomes verifiable on request.
Two engagements. Two stacks. One architecture, proven twice.
- 0
- Live engagements
- 0
- Architectural kernels built
- 0
- Bounded contexts modelled
- 0
- Legacy lines retired
The common kernel
Both engagements run on the same hand-built kernel.
Commands change state, queries read it, and domain events let the rest of the system react — over a CQRS message bus. Around the domain, ports and adapters keep it pure. The same kernel is reused by every bounded context in both case studies.
CQRS message bus live
Ports & adapters live
Case studies
Each project is the same idea, tested differently.
Working together
Your system could be the next proof.
If you're rebuilding a critical part of your operation and want architecture that compounds instead of decaying, I'd like to hear about it.