BayOS
AI-powered shop management layer for independent collision shops. BayOS sequences active jobs, tracks parts readiness, and surfaces today's optimal work order — giving managers and techs the best action at every moment. Picks up where DentStream leaves off.
Problem Statement
The Problem
An idle tech at 9am is a car that won't ship on time.
Independent collision shops are built around the bay, not a system. The day starts with verbal assignments, half-guessed parts status, and techs grabbing whatever looks ready. There's no sequencing engine telling anyone what to touch next — just instinct, habit, and whoever spoke loudest at the morning huddle. The shops know throughput is leaking. They just can't see where.
Existing tools don't solve this. CCC One and Tekmetric are record systems — they log what's in the shop, not what to do next. They tell you a car exists. They don't tell you which tech should touch it, whether the parts are confirmed, or what happens to three other jobs if you pull someone off a tear-down. That decision still lives in one person's head.
For shops on Direct Repair Programs, the cost is compounded. Insurers monitor cycle times obsessively — because every extra day a car sits is another day of rental they're covering. Loaners run out. Customers get stranded. When a shop consistently runs over, insurers pull them from the program entirely. Most shops manage DRP commitments the same way they manage everything else: by memory. BayOS was built for the shops that can't afford that anymore.
Approach
What I'm Designing
Designing BayOS from the ground up — informed by direct observation inside working collision shops in Houston, TX. Currently in active design. Build phase to follow.
Outputs today's optimal work order given active jobs, technician capacity, repair stages, and parts readiness. No verbal coordination required.
Tracks order status, ETAs, and core returns per job. Blocks scheduling until parts are confirmed in-house — eliminating duplicate orders and cascade delays.
Models the downstream effect of accepting a new job: which cars get pushed, realistic completion windows, and whether capacity can absorb it.
Parts cost, labor hours, insurance payout, and margin per job — P&L clarity that currently exists nowhere in the independent shop workflow.
Design Targets
Projected Impact
Design Thinking
Key Insights
- ◆The AI surfaces the best action. The manager still runs the shop.
Operational knowledge in most shops lives in one place: the manager's memory. BayOS doesn't replace that — it encodes it, reasons over it, and tells everyone the optimal next move. The manager gets leverage. The team gets clarity.
- ◆Record systems and decision systems are not the same thing.
CCC One tells you what's in the shop. It doesn't tell you what to do next. That gap is where throughput is lost — and where BayOS operates.
- ◆The wedge is sequencing, not scheduling.
Scheduling is a calendar. Sequencing is an optimizer. The question isn't 'when is this car due?' — it's 'which car should this tech touch next?' No existing tool answers that.
Next Project
BidOS →