How to Automate Trucking Operations: Turn Your SOPs Into Agents That Do the Work
If you want to automate your trucking operation, start with the SOPs you already have. Every carrier and brokerage runs on them, whether they're written in a binder or living in your best dispatcher's head. Pick one repeatable process, and let an AI agent run it 24/7 inside the system where your loads already live. Put a human in the loop to monitor and approve wherever you see fit. That's the whole method, and it's what Foundry does inside Alvys.
I'm not saying this as a software guy who read about trucking. I started as a dispatcher and grew an asset-based brokerage past $100M before Alvys existed. I've lived the night shift where three drivers go quiet and the paperwork keeps stacking. Every workflow I'm about to walk you through is one my own team used to do by hand, and every one of them is running in Foundry today.
What does it mean to automate a trucking SOP?
An SOP is a set of if-then decisions. "If a driver hasn't checked in within two hours, text him. Log what he says." You already know these cold. Most of them live in your head or a binder somewhere.
Automating an SOP means handing that loop to software. The agent watches your operation around the clock. The moment the condition hits, it does the step. Not a chatbot you have to prompt. A workflow that executes on its own, inside the system where your loads already live.
Which trucking operations can you automate today?
Here's what Foundry agents run right now.
Check calls. The agent spots any trip with no check call in the window you set. It texts the driver. It writes the response back into Alvys. Your dispatcher never picks up the phone for a load that's fine.
Detention. The agent scans every in-transit trip and watches dwell time at every stop. Hit your threshold, say every 50 minutes, and it sends the detention notice. If it’s disputed, the agent routes it to a human before any accessorial gets applied. You stop eating detention because someone forgot to start a timer.
Carrier outreach. The agent queries your open loads. It matches them against carriers you've actually used on that lane, with the MC number and rate already on file. Then it sends the outreach emails. Your team wakes up to replies instead of a cold list.
Paperwork into loads. Foundry provisions a dedicated email address for your docs. The agent classifies what comes in, pulls the fields, and builds the load straight from a PDF tender. Nobody retypes a rate con at 6 p.m. on a Friday.
Rerouting before trouble. The agent watches your live fleet against weather and risk data. It reroutes drivers before they hit the incident, not after.
There are more than 20 pre-built agents like these, covering things like rate audits, compliance, claims, and document verification before dispatch.
How do you build one? Four steps.
- Pick your most repetitive SOP. The one your team complains about. That's your first agent.
- Say it in plain English, or upload the SOP itself. Feed Foundry a PDF, a Word doc, or a bullet list. It assembles the workflow for you to approve. Prefer to start faster? Grab a pre-built template.
- Test it before freight moves. Run the agent on simulated data. No live loads. No risk. Prove it works first.
- Ship it small, then scale. Deploy on one tricky lane. Tell it something like "cover CHI to DAL for under $2/mile, only carriers with 98%+ on-time." When it earns your trust, run it across your whole book.
What keeps an agent from doing something dumb with my money?
This was my first question too. Freight runs on trust, and I wasn't going to ship anything that could burn it.
So every agent runs under Agent Shield. You set dollar thresholds per agent. High-impact actions pause and wait for a human sign-off, every time. Low-stakes busywork just runs. And every rate, route, and decision gets logged and traced. No black boxes. If you want to know why an agent did something, the record is right there.
Your data stays yours too. We hold zero-training agreements with every model provider, so your freight data never trains anyone's AI. On-prem options exist for companies that want it locked down further.
Do I need a tech department for this?
No. We send an engineer to sit with your team, learn your SOPs, and build the first workflows alongside you. You bring the freight knowledge. We bring the agents with you.
"Foundry is not AI for show. It is AI with a job description. For Spartan, it feels like having our own R&D team embedded inside the operation without having to build one ourselves. It helps remove the manual friction from freight so our people can stay focused on judgement, service, and elite execution,”
-Carlos M. Llanes Jr., Founder & CEO, Spartan Carrier Group.
The bottom line
You already wrote the playbook. It's your SOPs. Foundry just makes them run without you standing over them. Your people get their hours back for the work that actually needs a human: relationships, judgment, the hard calls.
More freight. Less friction. That's the whole point.
FAQ
What is Alvys Foundry? Foundry is the agentic workflow builder inside the Alvys TMS. It turns your standard operating procedures into automated AI agents that run around the clock: check calls, detention tracking, carrier outreach, document intake, and more.
Do I have to build workflows from scratch? No. Foundry offers 20+ pre-built agents for common carrier workflows. You can also upload an SOP document or describe the workflow in plain English, and Foundry assembles it for your approval.
Will an agent spend money or contact customers without approval? Not without your rules. Agent Shield sets spend limits per agent and pauses high-impact actions for human sign-off. Every decision is logged and traceable.
Is my freight data used to train AI models? No. Alvys holds zero-training agreements with every model provider. On-prem options are available for companies that want data kept fully in-house.
Can I test an agent before it touches live loads? Yes. Every agent can run on simulated data first. You prove it works before any freight moves.
