Augur ingests global open-source intelligence, scores it, matches it to the places you care about, and sends an alert before it reaches the headlines.
Raw intelligence enters on the left and leaves as a routed alert on the right. Every stage writes back to the same Postgres core.
Open feeds, webhooks and APIs
Adapters → canonical_event schema
Severity, geo, dedup, tagging
Watch zones turn events into relevance
Regional activity forecast and playbooks
Email, Slack, Telegram, webhook
Augur is glue, not a black box. Each capability is a small, inspectable service you can run, query and replace.
Every source, whether a feed, a webhook or a scraper, collapses into one canonical event: source, type, a 0 to 100 severity score, geo, payload and tags.
Define the places that matter: a refinery, a port, a route, a region. Augur flags every event that lands inside, with distance and a plain-language read.
When a watched zone is hit, Augur fires an alert by email, Slack, Telegram or webhook. Set per-zone severity thresholds so you only hear what matters.
Forecasts the volume of events to expect in a region or for a feed, so ops teams can pre-empt surges instead of reacting to them.
Run a what-if simulation on a situation before it happens: thousands of AI agents play out the scenario and surface the consensus outcome.
One schema, one database. Events, watch zones, forecasts and alerts are all queryable SQL, with no black boxes anywhere.
The same canonical pipeline powers a live risk dashboard, geofenced alerting, a crisis playbook studio and the event intelligence API.
A real-time view of global events relevant to your assets: an interactive map, a severity stream and per-zone hit counts. Built for ops, security and risk teams.
Get notified the moment something happens near a place you care about. Severity thresholds, quiet hours and per-zone routing keep the noise out.
Stress-test a situation before it happens. Scenario templates like a blocked strait or a sanctions shock are played out by autonomous agents into a consensus outcome.
Programmatic access to the normalized event stream, watch-zone matches and forecasts, so you can plug Augur into your own ops tooling.
Augur is built for research and decision support. The rules that keep it that way are enforced in code.
Augur produces operational awareness about events in the world. Every assessment is derived from the event's own fields and meant to be corroborated by a human before any consequential action is taken.
AGPL services stay isolated, secrets never hit version control, and feed terms of service stay the operator's responsibility.
.env.example templates onlyA live mock of the Crisis Playbook Simulator. The full simulator lives on Pro+ — this preview runs on a loop so you can feel the flow before you sign up.
The core pipeline is free to self-host forever. Managed tiers add hosting, the Signal API and simulation.
Kick the tyres on a single watch zone
Watch the places you care about, on your own
For analysts running real watch lists
For SOCs, ops centres and crisis desks
Augur ingests global open-source intelligence, normalizes every record into one canonical event, and flags the ones that happen inside the geographic zones you care about. Then it alerts you over email, Slack, Telegram or a webhook. It is operational awareness, not financial advice.
Ops, security, supply chain, insurance, maritime and aviation teams. Anyone who needs to know about geographic events near a set of assets, routes or regions, before the news cycle catches up.
It is a hybrid, self-host-first orchestration layer. Instead of forking upstream tools like Shadowbroker, the forecasting service or MiroFish, Augur unifies them through adapters, Docker, and a shared canonical data schema.
Shadowbroker and MiroFish run as separate containers via their own Compose files. They are never embedded into the MIT glue code, keeping licensing boundaries clean.
Every ingested record normalizes to: source, type, timestamp, severity (0 to 100), geo (lat/lon), payload and tags. One shape powers matching, forecasting and simulation downstream.
Yes. SHADOWBROKER_MOCK, MIROFISH_MOCK and TIMESFM_MOCK let you run the entire pipeline with mock data, which is ideal for demos and development before wiring real feeds and GPUs.
Spin up the pipeline in minutes with mock data, or talk to us about wiring real feeds and forecasts for your desk.