It answers carrier calls, runs your check calls, and writes the result straight into Alvys through the Alvys Public API, so the load record is updated before the call ends.
What it touches inside Alvys
Alvys is a good TMS. It still cannot dial a driver at 4 in the morning, answer the forty-third carrier asking about a posted load, or type the check call into the load while your coordinator is on another line. That gap is where status goes stale, carriers get voicemail, and covered loads slip. A voice agent sitting on top of your Alvys instance closes it without adding a seat to your team.
What it handles
Looks up the load the caller is asking about, confirms lane, equipment, and pickup window from the live record, and screens for the details your team needs before a human ever picks up.
Dials on your schedule or on a trigger, asks for location, loaded status, and ETA, then writes it back to the trip and stop so tracking is current without anyone typing.
Marks a stop arrived or departed from what the driver says on the phone, and can clear a bad timestamp when the driver corrects it.
Every call leaves a clean, timestamped note on the load, plus a full transcript and recording your team can pull up later.
Calls for the rate confirmation, BOL, or POD that has not come in, tells the driver where to send it, and flags the load once the document lands.
Rate negotiation, service failures, and anything unusual get warm-transferred to the right person with the load context already gathered.
Alvys Public API
Alvys ships a documented public REST API with OAuth 2.0 client-credentials auth, more than fifty endpoints, and webhook subscriptions signed with HMAC-SHA256. The agent authenticates as its own application, reads the live record mid-call, and writes the outcome back through the same endpoints your own team uses, under the permissions you grant it.
Endpoint groups we work with
Running a workflow that is not on this list?
How it works
You generate credentials in Alvys and grant the scopes you are comfortable with. Read-only first is completely fine.
We mirror how your brokerage actually runs: check-call cadence, who gets transferred, what makes a load a problem load.
Keep your existing number. Forward after hours, overflow, or a dedicated tracking line, and start outbound calls on your schedule.
Transcripts, recordings, outcomes, and the exact writes made to Alvys, all reviewable. You tune it from real calls in week one.
It uses the documented Alvys Public API. Authentication is OAuth 2.0 client credentials, the agent runs as its own application with the scopes you grant, and every read and write goes through supported endpoints. No browser automation, no shared logins, no scraping.
Both, if you allow it. The Alvys API supports updating loads, creating load notes, setting stop appointments, and recording stop arrival and departure, which is exactly what a check call produces. Plenty of brokers start read-only for a couple of weeks, listen to the calls, then turn writes on.
No. You keep your existing numbers and carrier. We take the calls you forward, whether that is after hours only, overflow when your lines are full, a dedicated carrier line, or everything.
It asks the way a human does, load number, PRO, or pickup city and date, then searches Alvys live and confirms the details back to the caller before doing anything else. If the match is ambiguous, it escalates instead of guessing.
It collects the load, the carrier, the MC, and the number they want, then transfers to your rep with all of that already gathered, or takes the offer and posts it as a note for a callback. We do not let it commit to a rate unless you explicitly define the rules it can work inside.
Most brokerages are live in about a week once we have API credentials and your call workflows. Alvys is one of the easier ones because the API is public, documented, and covers the objects a voice agent needs.
The main page: what a voice agent covers across carrier sales, check calls, and after-hours dispatch.
Read moreHow outbound check calls run themselves and land back in your TMS as timestamped status updates.
Read moreAnswer every is-this-load-still-available call, qualify the carrier, and hand off warm covers.
Read moreAPIs, webhooks, MCP, and what we do when your TMS does not expose the endpoint we need.
Read moreBook 15 minutes and we will walk through your check-call cadence, your carrier call volume, and exactly which Alvys endpoints the agent would touch. Flat monthly pricing, no per-seat fees.
Part of our voice AI for freight work, alongside AI voice agents for every other kind of business.