Multi-tenant by design
Each client is a fully separated workspace — its own taxonomy, its own data residency, its own access controls. Analysts switch context without leaking signal between books.
Use case · Agencies
Multi-tenant by design, brandable per client, and structured so each engagement has its own taxonomy, queue, and report. Media, PR, OSINT, and intelligence agencies use mediaIntel to deliver to clients without standing up their own pipeline.
What the engine does for agencies
Each client gets their own taxonomy, queue, and report template — without the agency having to operate a separate pipeline per engagement. You pick the policy, we run the model.
Each client is a fully separated workspace — its own taxonomy, its own data residency, its own access controls. Analysts switch context without leaking signal between books.
Reports, alerts, and dashboards ship in the agency's colours and voice. Customer-facing artifacts carry your brand, not ours. Client never has to know the engine underneath.
Each client's pattern library is bespoke — luxury brand watchlist, regulator narrative tracker, ORC pattern leaf, sovereign-info monitor. We map your policy doc to detectors during onboarding, with the right tool picked for each step in the pipeline.
How it fits the agency workflow
Agency margins live in analyst hours. The engine takes the bulk-classification and triage work off the queue so analysts can spend their time on synthesis, narrative, and the part of the deliverable a model can't write.
Triage and classification done in-engine; the analyst opens the queue at the synthesis step, not the labelling step.
Every event logs the model version, the analyst action, and the timing. Exportable to whichever billing or case-management system the agency uses.
Per-tenant configuration covers where the pipeline runs (managed cloud, in-country, on-prem) and which tools are used at each step — so a client's procurement preferences don't force a re-platform.
Our promise to agency partners
We are building the multi-tenant deployment surface — per-client workspaces, per-client taxonomies, per-client export templates — that lets a media or intelligence agency add a new engagement without re-platforming. We are pre-launch on the agency tier; these commitments are how we will structure partner pilots, not metrics from a running book.
One sprint
Target onboarding from contract to first client report
Tenant-isolated
Separate taxonomy, residency, and access controls per client
White-label
Client-facing deliverables carry your brand, not ours
Residency choice
Managed cloud, in-country, or on-prem — per tenant, per client
From conversations with PR shops, media monitoring teams, and OSINT firms scoping a partner tier.
Yes — client-facing artifacts (reports, alerts, dashboards) ship in your colours and voice. The agency owns the customer relationship; we don't sell direct to your client books.
Each client is a fully separated workspace — its own taxonomy, its own data, its own residency choice, its own access controls. Analyst accounts can carry multiple workspace roles, but signal does not leak between them.
Yes — residency and the per-step pipeline configuration are per-tenant. A client that requires self-hosted in their own data centre, or in-country managed cloud, is a workspace setting — not a separate product SKU. Pilot conversations cover which posture fits the client's procurement and compliance constraints.
Partner-tier pricing is per active tenant + usage, with volume discounts beyond the first few workspaces. Pilot terms cover the first client engagement end-to-end at flat fee; we work out the structured terms after the first delivery.
No. The partner motion is structured so we route direct inbound to the agency in the relevant vertical or region. We don't run an analyst desk and we don't take direct client work that would compete with a partner book.
Agencies — partner tier
Tell us the first engagement you'd run the pipeline against. We'll stand up the tenant, the taxonomy, and the report template inside a sprint — and you own the customer relationship end-to-end.