Operational Intelligence
The answer lives between the systems.
A semantic layer, decision workflows and governed agents that read across all four pillars — and act, under approval and full audit. This is the pillar that makes the other four worth having together rather than separately.
What it does
The core capabilities, live in the product today.
Cross-pillar correlation
A question spanning video, identity and data becomes one plan, authorized leg by leg in the pillar that owns each dataset. The answer shows the plan, the queries it generated and what was withheld.
Each leg of the plan is authorized where its data lives, so the answer is bounded by the asker’s own grants in each pillar — and the withheld list says what you were not allowed to see rather than quietly returning less.
A semantic layer over all five
Metrics defined once, with certification state, so two dashboards cannot disagree about what revenue means.
A metric is defined once and carries its certification state, so the number on one dashboard is the number on every dashboard — and an uncertified definition is visibly uncertified, not quietly different.
How it works
- Ask
Ask one question across everything
A question spanning video, identity and data becomes one plan, authorized step by step in the pillar that owns each dataset.
- Ask
See the plan before anything runs
Every answer shows its plan, the generated queries and the sources actually read — and what was withheld, and why.
- Console
Act under approval and audit
Decision workflows and governed agents act on the answer, and every action is approved and on the record.
The products in this pillar
Each one with its own page, its own published prices, and the same invoice.
See it in the console
Real screens from the products in this pillar, shown with demo data.
Ask: Ask it
A question about the north gate, answered with how it was read and the query it generated.
Workflows & Agents: A held step
An access review whose revocation is waiting for a person, with every step recorded.
Jobs it does
Use cases that run through this pillar, end to end.
- Offboarding everywhereWhen someone leaves, every kind of access ends together.
- Camera-to-sign-in investigationFrom a person on camera to the badge and sign-in records, in one case.
- Recovering from a bad changeUndo the bad change, whichever layer it was made in.
- Incident response with approvalsOne alert queue for everything, and a response that waits for a person where it should.
- Multi-site physical securityEvery site’s cameras, doors and sensors, watched from one place.
The ideas behind it
- One decision point for every permissionEvery service asks the same policy decision point whether an action is allowed, and every answer carries its reason. There is no second, quieter path to the data.Read the explainer
- A tamper-evident auditEvery allow and every deny, from every pillar, lands in one hash-chained record that can be verified, exported and kept under legal hold.Read the explainer
- Questions that cross systemsA question is split into one request per system that holds part of the answer, each authorized as the person asking, and the parts are joined on people, resources, time and place.Read the explainer
- Agents that stay accountableAn agent is a principal of its own with scopes and a budget. It acts only within the narrower of its own and its user’s permissions, destructive steps wait for a person, and every step is traced.Read the explainer
- How Platform Credits workEvery product is priced in one unit, Platform Credits, at a published rate per meter. Usage draws from one balance, and one invoice covers everything.Read the explainer
Where the work happens
Ask, in every console
The Ask surface: the plan before anything ran, the permission applied to each leg, the generated queries and the sources actually read.
Questions, answered
What can we actually ask?
A question that spans pillars — footage, access records, tables — becomes one plan, authorized leg by leg in the pillar that owns each dataset.
Does it only answer, or can it act?
Both. Decision workflows and governed agents act on an answer, and every action runs under approval and lands on the audit chain.
Why trust the answer?
Every answer arrives with its plan, the generated queries, the sources actually read — and what was withheld, and why.
What does it cost?
Usage at published rates on the pricing page, on the same invoice as the rest of the platform.
See it running.
Create a company and try it with your own data, or talk to us about your setup.