Dr. Leo — AI Field Operations Platform for HVAC

HVAC service companies run their field work on photos. A technician finishes a job, takes a picture, sends it in — and someone in the office has to look at every one of them to decide whether the work was actually done correctly.

Dr. Leo removes that bottleneck. Technicians submit job photos through Slack. The system runs image recognition over each submission, checks it against what the job required, and routes the result back — approved, flagged, or sent for review — without an office coordinator looking at every image.

What it does:

Technicians submit job photos and updates directly in Slack, so there is no new app to learn.

AWS Rekognition analyses each image; an AI layer interprets the result against the job type.

n8n orchestrates the whole flow — intake, analysis, decision, notification and logging.

Flagged jobs are escalated to a human; clean jobs pass through automatically.

Every submission is logged, so the operations team has a searchable record instead of a Slack scroll.

The point of the build was not the image model — it was the operational wiring around it: retries when an upload fails, error handling when a photo is unusable, and monitoring so the team knows the pipeline is alive.