How we work.

Human-AI synergy

Most research that uses AI treats it as an accessory: a person does the thinking and reaches for a model to speed up a task. We work the other way round. Analyst judgement and machine scale run as a single research process, aimed at problems neither could handle alone.

The AI partner is given explicit authority to do three things: recommend a direction, draft the work, and challenge the analyst's reasoning. The analyst reviews every recommendation, edits every draft, and has to answer every challenge. The decision, and the accountability for it, stay with a named human.

Why it suits the work

The ecosystems we map (open-source AI, anonymous inference, crypto rails, grey markets, and biological data) are too large, too fast, and too sparsely documented for conventional desk research to keep pace. They shift week to week, leave thin records, and reward no one for charting them. Synergy is what lets a small, independent team map them at all, and keep the map current.

The same capability produces everything we publish and sell, from regulatory briefings to subscription intelligence to training. One research engine feeds all of them.

The human decides

Scale without judgement is just volume. Every finding we publish is owned by a named analyst who can defend it, and nothing reaches a regulator, a client, or the public without that review. The challenge runs both ways: the AI is expected to push back on the analyst, and the analyst on the AI. All of it serves one end: getting the finding right.

Independence is part of the method

Independence runs through the research itself. It shapes what we take, what we sell, and what we refuse to do, and it is what makes our findings worth citing. We set it out in full, as commitments you can hold us to.