01 / Investigations

Work in public.

Current systems, experiments, and observations from the lab.

Systems

Shared memory, agent workflows, approvals, reporting, and creative operations.

Agents, handoffs, and workflow state for repeatable creative operations.

Briefs turned into reviewed, publishable outputs.

A shared tool layer for model-driven work.

Campaign intelligence, research systems, living brand knowledge, and content engines.

Signals, search, and model output held in the same decision frame.

Long source material shaped for reliable model use.

Policy research with traceable claims and citations.

Client-facing AI products, interactive brand systems, and AI-native campaign surfaces.

A toyetic canon and AI-native media systems experiment.

A live, reactive talk surface for forty slides and audience participation.

An interactive demonstration of text layout without the DOM.

Evaluation, review systems, safety checks, drift detection, and proof attached to the work.

A benchmark for judgment, coherence, voice, and refinement.

Longitudinal evaluation for high-stakes AI workflows.

Small experiments that measure, keep, revert, and repeat.

running2026-05

We build these systems for creative organizations, too.