Insurance-grade evidence, signed, sourced, timestamped.
Insurance runs on evidence. When a claim is contested, or an AI or robot decision is later questioned, what matters is simple: what was true, and what did the system act on, at the time? Dynamic Feed turns the live conditions and data it serves into Ed25519-signed, provenance-stamped, archived records, captured as they happen, so either side can re-check them later without trusting your word or ours.
x7Qa…verify against /.well-known/keys
After the fact, the inputs are un-provable.
A claim turns on what was true at a specific time and place, the weather, the flood level, the sea state, the hazard, the position. Most of that arrives as self-reported logs or screenshots pulled together days later. None of it is independent, and any of it can be questioned.
There is now a harder version of the same problem. When an AI or autonomous system acts and something goes wrong, the question becomes what data did it act on, and when? Underwriters increasingly treat that as too much of a black box to price: through 2026 standard carriers have been adding explicit AI exclusions (the Verisk/ISO generative-AI liability exclusions took effect 1 January 2026), while specialist insurers such as Armilla and Munich Re move into the gap. Whether it is a contested claim or new AI cover, the same artefact makes the risk assessable: an independent, third-party record of the inputs, captured at the moment they existed, signed and timestamped, and re-checkable by either party.
A signed record of the inputs anyone can check.
The conditions a claim argues over, and the data an AI acted on, delivered the way every datapoint we publish is, signed, sourced, and archived.
- Live conditions for claims. Weather, flood, hazards, marine and GPS integrity, signed records of what was true at a time and place.
- The data your AI acted on. Every datapoint carries a provenance envelope, source, licence, exact measurement time, age and an Ed25519 signature, so you can show precisely what data we served and how fresh it was.
- A signed decision trail for embodied AI. The awareness API returns a go / caution / no-go verdict for a robot at a location, Ed25519-signed, a contemporaneous record of what the system was told before it acted.
- RFC 3161 proof-of-existence. A record's fingerprint is time-stamped by an independent RFC 3161 authority, instant, no wallet, independent evidence it existed, unchanged, no later than a given moment.
- Append-only archive. Records are written as they happen to an append-only signed store, so they cannot be quietly rewritten after an incident.
- Verify it yourself. A public verify page and the open-source verifier let either party re-check a signature independently, in their own code, without trusting us.
Don't take our word, verify a live record.
This fetches a live signed record from Dynamic Feed and verifies its Ed25519 signature in your browser, against the published key. Then flip one byte and watch verification fail, tamper-evidence demonstrated live, not asserted.
A valid signature shows Dynamic Feed reported this exact record at this time, and that it has not been altered since. Change one byte and verification fails, and anyone can check. It does not prove the underlying value is objectively true, or that acting on it was the right call. Provenance, not omniscience.
Built for the people who carry the risk.
The witness your claims engine cannot be
When an algorithm prices and pays a claim in seconds, what has to hold up later is the record of the data it acted on. An automated claims engine cannot witness itself. We give you and your reinsurer an independent, signed record of the reading behind every automated decision.
Your input-data integrity layer
A verifiable record of the external data we served you, when, from where, and how fresh, so you can show what Dynamic Feed served you at a given moment. A defence artefact if a decision is later questioned.
An independent risk-control signal
A third-party, tamper-evident record of the provenance and freshness of the Dynamic Feed data an applicant relied on, which you can require and re-check independently.
Reconstruct the conditions
Signed, timestamped weather, flood, hazard, marine and position records for the exact time and place in question, captured as it happened, re-checkable by either side.
A signed trigger receipt
A signed, contemporaneous record of the reading behind a parametric cover, and a neutral determination of whether the trigger was met, each party can re-compute. For capital-markets desks, see capital markets. We do not calculate risk, settle, or move any payout.
Evidence for CPS 230 & ADM transparency
Signed, non-PII, provenance-rich feeds can serve as one piece of verifiable evidence to support AI-supply-chain assurance under APRA's CPS 230, and ahead of the automated-decision transparency obligations commencing 10 December 2026. See AI governance.
One call → a signed record.
Call the feed
One keyless MCP or REST call for the conditions or the awareness verdict you need. No key to get started.
Get the inputs
The response carries the reading or verdict, each with its source, licence and exact measurement time.
Keep the record
Every response is Ed25519-signed and written to the append-only archive, so you can produce it months later.
Verify independently
Either party re-checks the signature on the public verify page or with the open-source verifier.
One price, the whole platform.
Insurance evidence is part of Dynamic Feed, not a separate bill. Start free and keyless; paid tiers (Builder, Scale, Enterprise) are in early access — request access and tell us your workload. The same signed, sourced, archived records on every tier.
Insurance evidence is one application of the broader evidence layer. Every tier is self-serve, even Enterprise. Mint a key and you are live in seconds. Want the parametric, capital-markets angle? See capital markets. Need weather conditions at a site on a date, signed for a claim? See weather evidence. Defending an automated underwriting or claims decision? See AI decision evidence. Every input is a verified fact.
Evidence, not a certification.
Dynamic Feed is an independent evidence layer, not an insurer, not a broker, and not financial, insurance or legal advice. We never price risk, trigger a payout, or recommend cover. Signing and RFC 3161 timestamping let anyone verify a datapoint existed at a specific time and has not been altered since, they do not prove it is objectively true, nor that acting on it was correct. Many readings are model forecasts, not measurements. This is tamper-evident, not tamper-proof, and not a safety or compliance certification of anyone's AI. Treat it as one independent, verifiable input, never the sole basis for a coverage, claims, or safety decision.
Fully automated. Run by AI, end to end.
No sales calls and no waiting. Mint a key and you are live in seconds, on any tier of insurance evidence.