Every repair order, line item, technician, and customer interaction your shop produces is a signal. CSA structures those signals into patterns you can read — what closes at what price, which jobs come back, which parts fit which makes — and compares them to the rest of the network.
Every line item is auto-categorized by what was actually done, not by the words on the invoice. Search by intent, group by category, compare across the network.
Brake jobs land in brakes. Cooling repairs land in cooling. Diagnostics are diagnostics, not "shop supplies." Closed ROs are reclassified as the system learns.
Every labor item carries a fit-to-vehicle percentage drawn from your shop’s history and the network’s. Verified means another shop has done it on this exact make.
Pair-this-with-that suggestions derived from co-occurrence in real ROs — not a vendor’s template, not someone’s opinion.
A customer returning within 30 days for a related issue is one of the clearest signals of work quality. CSA computes comeback rate per category, per technician, per make — and surfaces the repeat lines so you can act on them.
Brake work, cooling, suspension, electrical — with each technician’s rate relative to the shop and the network. Coaching signal, not a scoreboard.
Top recurring failure patterns surfaced as a list, with the vehicles and miles between visits. Catch a parts-quality drift or a procedural gap before it spreads.
From the side panel, ask in plain words. Otto reads your data, draws on the network where it helps, and answers in a table or a short summary — with the rows he’s referencing.
Open a job: the labor (your shop’s average), the parts (your usual), the tech with availability are already on the page.
Ask Otto a whole-shop question. He answers with a table, the rows he’s referencing, and the action you can take next.
Comeback rate, margin by category, technician throughput — the numbers the manager already tracks, sourced from the same model.
Private alpha, in production. Partner-reviewer slots open in waves.