Method

Across every project, the work moves through three stages. They are not strict phases — research keeps running long after the first cameras roll; format is reconsidered as the research deepens; and the after-life of a production is designed in from the start. What follows is the shape, in the order the work tends to settle into. The same pattern applies whether the deliverable is a documentary series, a traveling exhibit, an XR piece, or a long-form podcast.

Story shape emerges from the material. It is never imposed on it.

We work from a structured research record — every person, organization, place, event, source, and story lead the platform has documented, cross-linked, and anchored to a citation. Each project draws on the record it inherits and contributes back to the one the next project will inherit.

The methodology is AI-supported, editorially governed. Custom production tooling accelerates ingest, surfaces cross-references across hundreds of entities, and flags patterns a producer working from raw notes would miss. The editorial threshold — what counts as a primary source, what is worth pursuing, what gets cited and how — is set and held by the producer. The tooling is the line. The judgment is not.

This is unusual infrastructure for a documentary platform, and it is the asset most worth understanding. Research compounds across projects; every claim is sourced; the platform builds field infrastructure, not just media.

The ethics work is most visible in our immersive media, where the risk of extraction is highest. Four commitments codified through the platform's current XR work apply to any immersive piece we produce — and a layer of ethics-partner governance sits underneath all four: a university bioethics center, a community IRB, and direct working relationships with the people whose lives the work inhabits.

• Disclosed consent. Anything the piece will measure, sense, or use is disclosed up front, in plain language. Consent is opt-in.
• Personal biometric baselines, not population norms. Calibration is to the participant's own physiology. Biometrics are a timing signal, never a leverage signal.
• Low-stakes, reversible asks. When the work invites action at heightened attention, the ask is small, civic, and reversible.
• Refusal of empathy theater. The work refuses to manufacture catharsis. Recognition, accountability, and structural understanding are the goals. ::

04.

Community-led, in practice

The phrase "community-led" gets used widely enough to mean very little. What we mean by it is specific — four working practices that govern how a project relates to the people whose lives it inhabits. The practices are structural commitments, not a sensibility; none of them is optional, and together they distinguish community-led work from documentary that merely consults a community along the way.

• Subjects are partners, not sources. Relationships are formalized before production begins. Consented participation cannot be retrofitted.
• We co-produce. We don't extract. A partnership starts with a conversation about a community's constituents and their interests — not with a pitch about what we want to make.
• Story shape emerges from the material. A community's lived experience, leadership, and existing organizing work shape the structure of the work.
• Archive lives with the field. Documentation, interviews, research, and primary source material are designed to be useful to partner organizations after the production ships. ::

This stance has costs. It is slower than the alternative, and it narrows the projects we can responsibly take on. The costs are load-bearing.