Mimir · Exposure intelligence
If you read it on Bloomberg,you're already late.
Mimir reads ships, planes, satellites, and ground sensors, then traces what they mean to the companies you actually hold. Every hop names its mechanism and carries its receipt.
Four feeds, one priority queue
The signals that move stocks before the print does.
Revenue follows physical activity by hours or days. Mimir watches the activity.
| Source | Refresh | Ticker map | Fail mode |
|---|---|---|---|
| AIS vessel positions | hourly rollup | shippers, oil, container, dry-bulk | partial visible |
| ADS-B aviation (adsb.lol) | ~60 seconds | airlines, lessors, MRO | fail-closed |
| USGS earthquakes | ~15 minutes | fab regions, insurers, utilities | fail-closed |
| NASA FIRMS fires | hourly | utilities, insurers, ag | token-gated |
| GDELT events | hourly | country → sector exposure | partial visible |
| N2YO satellites | hourly cache | defense, satcom | CelesTrak fallback |
| NOAA space weather | ~hourly | satcom, grid, aviation | fail-closed |
“The print lags the physical signal by hours. Sometimes days. Mimir watches the source, not the wire.”
Why this exists
How a signal lands
One event, traced from sensor to ticker.
A Bab-el-Mandeb slowdown is not a headline. It's a vessel-count drop, then a transit-time spike, then a curated ticker fire. Mimir runs all three steps before the news cycle gets to step one.
- T + 0hAIS rollup detects anomaly
Vessel count through Bab-el-Mandeb drops 38% below 14-day baseline. Northbound transits down sharper than southbound. Z-score 2.7.
- T + 0hCurated map fires
The Bab-el-Mandeb → Red Sea routing dictionary maps the anomaly to specific tickers. Not a string match on "shipping" but a hand-built list of who actually transits this strait.
Bab-el-Mandeb slowdown→Red Sea reroute via Cape adds ~10 days→ZIM · MAERSK-B · HAFNIA · EXPD - T + ~3hSIGINT rail re-sorts
If your watchlist holds ZIM, this signal jumps to position #1. If you don't, it stays mid-rail. Same global feed, your priority order.
- T + 24hThe wire catches up
Reuters publishes "Houthi attacks disrupt Red Sea shipping." ZIM gaps up. You've been watching for a day.
The exposure graph
That path is one edge in a bigger graph.
Under the ranked feed sits a browsable ontology: sources connect to the mechanisms they trip, and mechanisms connect to the companies they move. Walk it in either direction, and every edge carries its receipt: the basis, the source, and the date it was asserted.
Same intelligence, on a live map.
Browse instead of querying. Atlas is the map mode of Mimir: same feeds, same composite keys, different question. "What is happening, where" instead of "what is exposed in my book." Click any point on the map and you land back in Mimir with that signal pre-selected. And when separate anomalies line up, Atlas clusters them into a Situation, with the evidence trail that formed it.
vs the alternatives
Three places retail traders look first. None of them watch the source.
| Twitter / X | Bloomberg headline ticker | News terminals (RTRS, Benzinga) | Mimir | |
|---|---|---|---|---|
| Source of the signal | Whoever posts first | Reuters / wire desks | Wire feed | Sensors / sats / AIS |
| Median latency to event | Highly variable | Hours | ~Minutes after wire | ~minutes to ~1h, by feed |
| Ranked by your book | No | No | Watchlist filter only | Exposure-weighted |
| Mechanism cited | No | Buried in body | No | 3-hop chain, editable |
| Fail mode | Engagement bait | Black box | Silent drop | Fail-closed, visible |
| Annual cost | $0 – $96 | $31,980 | $1,200 – $9,000 | $348 (Pro, $29/mo) |
Competitor list prices retrieved 2026-07-08; Bloomberg list updated 2026-07-22.
Stop scrolling for the next disruption.
Mimir ranks it. Atlas shows you where. Both wired into your watchlist.
No portfolio, but a supplier and facility footprint? The same ledger runs against an operational book: supply chain risk.