Skip to content

Home/Threads

Coordination between agents

Separate parties have to act together without reading each other's insides.

Marketplaces, support centres and multi-agent systems share a structure: independent parties with private information must reach joint action through public signals alone. Fifteen years of building the human version informs the artificial one — a social layer that handles evidence, cooperation, governance and repair without ever inspecting private internal state.

Where it shows up

4 positions, earliest first.

  1. Junior In-House Entrepreneur · Rocket Internet

    Oct 2011 – Nov 2012

    • Set up a regional support centre, scaling it to 200 staff and USD 80k monthly revenue; led localisation of Zalora Philippines across marketing and operations.

  2. Co-founder · e-Cose

    Nov 2012 – Dec 2014

    • Built the business from plan to breakeven inside the first month; hired and trained the first 10 employees and drove 100% revenue growth over 12 months.

  3. Co-founder & Head of Marketing · Epigami

    Nov 2013 – Jan 2017

    • Grew a two-sided marketplace to 150k monthly organic visits and a pool of 5,000 tutors, outranking hundreds of local competitors; lifted conversion from 3% to 15% through iterative UX research.

  4. Independent Researcher · Relational AI & Cognitive Architecture

    2024 – present

    • ARCANA — Conceived and built a framework for modelling intelligence as relationships and evolving processes rather than isolated objects. The system represents how patterns form, stabilise, change and recur across agents and groups, with every transition reproducible and auditable.
    • Reasoning & collective intelligence — Unified perception and reasoning as geometry over relationships: structure defines possible connections, measurement establishes what evidence supports, dynamics shapes candidates and deliberation compares paths. A social layer separately handles causal evidence, cooperation, governance and repair without reading private internal states.