Scientist turned AI architect, building human-AI workflows that are useful, governed, and production-ready.
I work at the intersection of enterprise AI, data platforms, and real team operations. My focus is turning LLM potential into systems people actually trust and use.
My foundation is in chemistry and synthetic biology (PhD), so I naturally design AI systems like scientific workflows: hypothesis, instrumentation, validation, iteration.
That means:
- strong boundaries and clear system behavior
- measurable outcomes over hype
- human accountability built into the loop
- Human-AI workflow architecture for enterprise teams
- LLM + tool orchestration with governance by design
- Secure access to structured and real-time enterprise data
- Responsible AI boundaries, observability, and operational scale
- PromptGramming (live): rm2thaddeus.github.io/promptgramming
- Pixel Detective (live): rm2thaddeus.github.io/Pixel_Detective
- Aitor Skills (live): rm2thaddeus.github.io/Aitor_Skills
- Clearance: talk with my AI profile assistant
- Cortexus: custom GPT creation and governance
- AI-Thor: strategy, planning, and structured problem-solving
- Promptfessor: guided prompt coaching and iteration
- Codyssey: technical task execution in regulated contexts
- 20+ assistants deployed in business settings
- 400+ users enabled through AI workflow adoption
- 45+ hours saved per user/year through automation
- 2.5x productivity gains in technical services workflows
- 4,000+ annual interventions tracked across 11+ teams
- 226% YoY lead growth visibility from data-product instrumentation
- 500k EUR recovered in mislabeled returns