Fresh Authority Feed
New authorities, delivered the morning they file
A daily file of FMCSA operating-authority applications — brokers, carriers, forwarders — parsed from the official register, enriched with the federal census, and labeled honestly by stage. 1,574 applications parsed since June 4, 2026; latest register day June 11, 2026.
Sample
$0
- The full daily application file, delayed 7 days
- CSV + JSON, same schema as paid tiers
- Judge the data quality on real records before paying
Starter
$29/mo
- Day-zero daily application file (CSV + JSON)
- All authority types, all states
- Delivered by email every publication morning
- Cancel any time, keep delivered files
Pro
$79/mo
- Everything in Starter
- Census enrichment as records propagate (fleet size, contact fields, status)
- Granted-confirmation diffs — know when applications turn into authority
- BFF broker bond-cancellation notices (fraud screen)
- State and authority-type filters
Team
$199/mo
- Everything in Pro
- Static JSON endpoints + agent-callable MCP verification
- Multiple seats / delivery addresses
- Custom filter profiles
- Priority support from the people who run the pipeline
Month-to-month. Built on public-domain federal data; you are paying for parsing, enrichment, monitoring, and delivery — not for the data itself.
The honest comparison
Do it yourself vs. the feed
Every source we use is free and public. If you have engineering time, you can build this — here is what the build actually involves, from our own logs.
| Task | Doing it yourself | With the feed |
|---|---|---|
| Fetch the daily register | Call the motus.dot.gov report API each morning; signed download URLs expire in 15 minutes, so fetch and store immediately. | Done by 8 AM ET; file in your inbox. |
| Parse the PDF | Handle wrapped names, addresses that concatenate across lines, page footers that interleave mid-section, Canadian addresses, re-application name suffixes, and ~1% of rows whose fields extract out of order. | Already normalized; edge cases flagged per row. |
| Enrich with the census | Batch SoQL lookups against a 4.4-million-row dataset — and retry over following days, because most day-zero applicants aren't in it yet. | Enrichment lands automatically as records propagate (Pro). |
| Detect grants | Daily diff of docket appearance and status codes per tracked applicant, tolerant of multi-day census latency. | Granted-confirmation rows in your file (Pro). |
| Track bond cancellations | Poll the BFF report type (it publishes irregularly), parse single-page notices, normalize ligature artifacts. | Included and normalized (Pro). |
| Survive format changes | FMCSA's May 14, 2026 cutover to Motus broke pipelines built on the old endpoints. You own the next migration. | We own it — it's the product. |
Straight answers
Feed FAQ
Where does the data come from?
Two official U.S. government sources: the FMCSA daily register published at motus.dot.gov (operating-authority applications and BFF financial-security notices) and the FMCSA Company Census File on data.transportation.gov. Both are public-domain federal data. We fetch politely, parse, normalize, and deliver — we do not scrape private sites.
Is an application the same as having authority?
No. A register entry is a preliminary grant under 49 CFR Part 365. A 10-day protest window follows, the applicant must file insurance and BOC-3 evidence within 20 days of publication, and authority is effective only on issuance. The feed labels application-stage and granted-stage records separately, and our granted-confirmation diffs tell you when status actually changes.
How fresh is "day zero"?
The register posts at about 7:00 AM Eastern each publication day. We parse it the same morning and your file reflects that day’s filings. The free sample tier is the identical file, delayed 7 days, so you can judge quality on real data before paying.
Why are many applicants missing from other databases at first?
New applicants take days to propagate into downstream datasets. When we joined three days of register filings (801 unique USDOT numbers, June 8–10, 2026) against the federal census file, only 46.4% had appeared in it yet. The register is where those records appear first, so parsing it day-zero is how the feed surfaces them before they show up in downstream databases — that is the point of the feed.
What format is delivery?
CSV and JSON, delivered daily by email link on Starter and by email plus static JSON endpoints on Pro and Team. Team adds the agent-callable verification endpoint (MCP) so your own tooling or AI agents can query records directly.
Can I cancel any time?
Yes. Month-to-month, cancel whenever, and you keep every file already delivered.
Could I build this myself?
Yes — the sources are free and public, and we list exactly what the build involves in the comparison table on this page. The feed exists for people who would rather spend that engineering time selling.
Not ready to subscribe?
Take the weekly note instead — filing volume and data changes, one email, no pressure.