Folio

Folio is research, not a finished product. The question it asks: if the work product of a creative practitioner is mostly intangible (strategies written, decisions made, opinions held, engagements under NDA), what's the right surface for that work to live on for visitors? A LinkedIn profile? A portfolio of static case studies? Or an AI agent that draws on the practitioner's full history and lets visitors interact with the way they actually think?

The hypothesis: the static portfolio is a 1990s paradigm, practitioner-as-document. The next paradigm is practitioner-as-agent: a knowledge layer the visitor can ask questions of, get pushback from, and probe across topics rather than scroll through a flat document.

We're working through the conceptual and technical questions a public-facing practitioner-agent has to answer. Voice fidelity: does it actually sound like the person across registers, not just on prepared topics? Refusal handling: what does it not speak to, and how does it refuse without breaking the illusion? Accuracy on history: does it claim only what's true, with evidence the visitor can probe? Conversational boundaries: when does it hand off to a human, and how cleanly? None of these have settled answers yet. That's what makes it research.

The work is iterative. Each pass refines the data model for a practitioner's knowledge, the interaction surface for visitors, and the agent's voice and conversation policy. The project moves when the conceptual questions move, not on a fixed schedule.

Why ask the question. Static portfolios optimize for the visitor who skims for thirty seconds. An agent-based portfolio optimizes for the visitor who actually wants to evaluate fit: the one who'd ask a real practitioner five sharp questions over coffee. The conversion economics are completely different.

Why not ship yet. A public AI agent that represents you is a high-stakes interface. Voice fidelity, refusal handling, accuracy on history, defensible conversational boundaries. All need to be solved before the first visitor types. Shipping early would ship wrong. The right path is to keep iterating until the answers are sharp enough to defend in a real conversation.

Where it goes. The OUTURE site you're reading is the static-first version of the same problem. A future iteration may include an agent layer that lets visitors interact with how OUTURE actually thinks, instead of just reading what we wrote down. Folio is where we're working that out.

Internal research, in development. The thread is open and continues; the questions it raises are still the right questions for the post-static-portfolio era.