The problem I wanted to solve
AI can change code quickly. That is not the same as knowing what changed, why it changed, whether it passed, and when a human should stop it.
Software developer and systems specialist building practical AI products.
Years in support taught me where work actually breaks. Now I turn that experience into practical AI products with clear plans, human approval, and proof that the system did what it said.
Amsterdam, NL · Open to AI and product engineering roles
Scroll to inspect the proofsystems → product → AI
One continuous chain of evidence
- Human intent
The person defines the outcome.
- Plan
A reviewable plan is frozen first.
- Build
Work happens in an isolated Git worktree.
- Verify
Checks and independent review leave receipts.
Ares is my governed AI software delivery workspace.
Working pre-beta app · native Chat and Work proven · packet Work acceptance still open
So I made authority and evidence part of the workflow.
Ares freezes the approved plan, isolates each change, and requires independent review. Parallel work is allowed only for disjoint write sets, and every merge still belongs to the owner.
Ares evidence · 23–25 Sep 2026
- harness checks passed on the 23 Sep build
- tests in that complete suite
- median tokens across five reviewed Work runs
What the architecture protects
Human authority
Versioned plans require approval, and Ares never chooses Merge for the owner.
Traceable work
A reviewed change can be re-gated after the base moves. One live merge succeeded, but its receipt still included unrelated history in the diff hash; that defect remains open.
Private evidence memory
The Episode writer is now built and tested to record each merged execution separately. Reusable patterns still need evidence from separate Objectives and owner approval.
Local-first context, with limits
Packet Chat was accepted. Five packet Work edits passed review and verification at a 15,087-token median, but the merge-rate comparison and Work acceptance remain open.
Notes from the workbench.
I learn by building. These are short records of what I tried, why I tried it, what failed, and what I am taking into the next iteration.
Ares never gets to choose MergeAI agent engineeringOpen note
Human in the loop means little if the system can approve its own outcome. Ares can propose, build, review, and verify, but only the owner can integrate the result. That boundary is now part of the real execution path.
A successful merge still needs an honest receiptDaily-use reliabilityOpen note
Ares re-gated a one-line change after the base moved, and I merged it in the native app. The merge receipt still measured its diff from the old base, pulling unrelated changes into the hash. The merge worked; that evidence bug is still open.
Five Work runs met the token targetLocal-first context routingOpen note
Five real documentation and test edits passed verification and independent review on the first attempt. Median provider use was 15,087 tokens against a 25,000-token target. That meets the cost half of the study, not its full acceptance bar.
Systems experience is my AI advantage.
I know the distance between a promising demo and a tool people can rely on. My background is support, administration, documentation, migrations, and translating between technical systems and the people using them.
- Identified blockers in process flows, proposed improvements, implemented changes, and kept documentation current.
- Led the migration upgrade from a discontinued Excel version to Microsoft 365.
- Contributed to the innovation team and led the creation of the game room.
- Acted as the first point of contact between business partner Trip.com and internal employees.
System administration and support at the Ministry of Finance in Suriname, SoftTech, Change=, and igen placements at Sanquin and Rijkswaterstaat. I learned to diagnose clearly, communicate calmly, and improve systems after incidents.
Education
NATINMBO level 4 · ICT Application Development · completed 2018, diploma issued Feb 2019
Inholland University of Applied SciencesBusiness IT & Management coursework · 2022 to 2024 · no degree claimed
Languages
Dutch and English · native proficiency / Italian · currently learning
Bring me the complicated part.
I am looking for a team where I can combine systems thinking, product curiosity, and practical AI engineering, especially close to real users and real operational problems.