I build at the frontier where AI, complex-systems thinking and venture craft meet. My bias is Deutschian: problems are soluble when we create the right explanations—so I design arenas where ideas can collide, mutate and be tested fast.
Over a fifteen-year, six-country path I've been bridging builder and theorist. I began in large-scale data and AI engineering, then co-founded Lucidminds AI—initially a responsible-AI firm offering agent-based modelling and simulations-as-a-service. It soon grew into an R & D collective and ultimately a venture studio shipping products that based on new economic mechanisms in emergent markets. Along the way I have mapped incentive structures, designed coordination protocols, and shipped infrastructure and products that serve thousands of users while remaining flexible enough for rapid iteration.
If your challenge sits at the edge of current knowledge and needs both conceptual rigour and hands-on execution, let's talk.
Self-organizing knowledge fabric that treats research as conversation. Chat-native archiving meets graph-native discovery, where fragments of thought—quotes, sketches, sudden clarity in conversation—become building blocks of deeper inquiry. The system learns your research ontology through lightweight training, maintains a quiet sense of directedness across projects, and surfaces connections when the season is right. Not another search tool, but a companion mind that makes intellectual drift continuous and returning inevitable.
Core innovation: Moving beyond retrieval efficiency toward "wondered knowledge"—systems that tend to questions over time rather than delivering instant answers. Built with modern LLM fine-tuning, embedding architectures, and production deployment patterns.
ABM framework to simulate and finance nature-based interventions; evolved from TreesAI. Role: translated mechanism design into framework & pilots; co-authored paper; led early fundraising.
Compute-to-data stack: run analytics/ML where data lives to unlock sensitive datasets without data egress. Role: system design, policy→technical controls, prototype pipelines.
Socio-technical design templates for collective intelligence (transparency, privacy, human oversight), later applied in TreesAI/GUS/Dataville.
Books I've read that left traces