Trilix AI
Case study · Quick lube

An AI Front Desk for a Multi-Store Quick Lube in Texas

An AI phone agent on the front desk, a second AI checking its work, cameras in the stores and a fair way to coach the team.

We keep client names private. This page covers what we built for this operator and how it runs.

The situation

This operator runs several quick lube stores in Texas. Like every quick lube, the phones ring all day.

Most calls are the same handful of questions. What time do you close? Can I just walk in? Can I book a time? Do you do state inspections?

Those calls land on the people who are also writing tickets and talking to customers at the counter. The goal was for every caller to get a good answer without pulling staff away from the cars.

The owner also wanted a clearer view of what happens inside each store, and a better way to coach employees on how they talk to customers.

What we built

An AI front-desk phone agent

The agent answers like a service advisor: cheerful, but professional. It handles store hours, bookings, walk-in questions and state inspection questions. When a caller needs a person, it transfers them to the front desk.

An audit agent that checks every few hours

A separate AI agent reviews the phone agent's calls every 2 hours during shop hours. When something looks off, it emails the founder. When everything looks fine, it stays quiet.

AI camera analytics

Camera analytics are rolling out store by store. Eight stores are done and three are left.

Conversation intelligence for coaching

A platform grades employees' conversations with customers, so managers have something specific to coach from instead of a general feeling.

How it works day to day

A customer calls a store. The AI picks up, answers the question or books the visit, and passes the call to the front desk when it should. The team stays with the customers and cars in front of them.

Every 2 hours during shop hours, the audit agent goes through the latest calls. Most of the time nobody hears from it. If it finds a problem, the founder gets an email.

That audit has already earned its place. It showed that the smaller AI model we started with was ignoring some of the booking rules. So we upgraded the phone agent to a stronger model. We'd much rather catch something like that in an audit than hear about it from a customer.

The coaching platform needed care, because nobody wants to feel watched. So we built privacy into it from the start. The raw recordings are visible only to the employee in them. Management sees the graded reports.

That way managers get what they need to coach, and the recordings themselves stay with the person they belong to. We also drafted the employee agreements for the program around Texas law.

The camera rollout runs alongside all of this. Each store gets set up in turn, so the owner isn't dealing with every location at once.

What's next

We're now working on a marketing audit. It compares marketing spend to new customers, month by month. The goal is for the owner to see how marketing spend lines up with the new customers coming through the door.

The last three stores on the camera rollout are also on the list.

If you run quick lube stores, our quick lube page shows the full system, and the AI phone agent page goes deeper on the calls.

Want something like this for your stores?

Tell us how your stores run today. We'll show you what we'd build first.