This is not a software license. It is not a platform subscription. And it is absolutely not another login the shop's estimator has to remember. Collision Intelligence operates on a hybrid managed-service model: James Bonaguro configures the Routines, tunes the triggers, monitors the outputs, and keeps the integration warm with CCC, Mitchell, ARMS, and the OEM portals. Anthropic's infrastructure runs the compute. The shop does exactly one thing — read the results. Agent outputs land in Slack, email, or a mobile text message. Supplements land back in CCC ONE with the justification lines pre-populated. The estimator reviews and approves. The operator sees the ROI report on the first of the month. This page lays out the three pricing tiers, the pilot program, and the four-week implementation roadmap from handshake to live revenue capture.
The collision industry has been burned by SaaS. Another dashboard, another seat license, another implementation consultant. The managed model is the opposite: the shop never touches the machinery. Three parties. Three jobs. No handoff friction.
No dashboard rollout. No user provisioning. No training week. The shop's SMS stays the system of record — agents integrate upstream and downstream of it.
Outputs route to the channels the shop already reads: Slack, email, SMS, or directly into the CCC ONE workfile as line items.
Every supplement line, every sublet reconciliation, every rental extension is surfaced for estimator approval via session URL. The agent proposes; the human confirms.
First business day of every month: a single PDF showing lines captured, hours saved, liability events documented, and dollars recovered. The report defends the invoice.
The shop pays for outcomes, not seats. If the agent does not produce, the agent is gone. The managed model keeps the incentives aligned.
Pricing is engineered to be transparent, outcome-anchored, and obviously profitable for the shop. Every tier clears its fee inside the first week of monthly capture based on the figures documented in the ROI Framework. Setup fees are one-time, non-recurring, and cover integration, trigger configuration, and the first two weeks of active tuning.
A single agent, deployed and monitored. Typically Agent 04 (E01 Predictor) — the flagship teardown defender.
Three agents of the client's choice, deployed and orchestrated together for compounding leverage.
All five agents, plus custom agent development as new pain points emerge. Built for MSOs and high-volume independents.
| Included | Reflex | Stack | Operator |
|---|---|---|---|
| Agents deployed | 1 | 3 | 5 + custom |
| 24/7 Anthropic cloud execution | ✓ | ✓ | ✓ |
| Human-in-the-loop session URL handoff | ✓ | ✓ | ✓ |
| Monthly ROI report | ✓ | ✓ | ✓ |
| Routine prompt updates | Quarterly | Monthly | Continuous |
| Strategy calls | — | Monthly | Weekly |
| Multi-system integration (CCC + Mitchell + ARMS) | — | ✓ | ✓ |
| Custom agent development | — | — | 1/quarter |
| Priority incident response | — | — | ✓ |
| Monthly fee | $650 | $1,750 | $3,200 |
| One-time setup | $1,500 | $3,500 | $6,000 |
| Expected payback | < 2 weeks | < 2 weeks | < 4 weeks |
All tiers billed monthly. Annual prepay available at 10% discount. Setup fees invoiced at contract signature; monthly billing begins on production go-live date.
The first three shops to sign establish the case studies that carry the next thirty. The pilot program is engineered to reduce the shop's risk to near zero and to produce published, verifiable proof within the first 30 days.
Four weeks from signed contract to live production revenue capture. No six-month consulting engagement. No training week. The roadmap below is the actual schedule — not a marketing timeline.
Ninety-minute discovery call with the operator. Audit the SMS (CCC ONE / Mitchell RepairCenter), the DRP mix, the sublet vendor roster, and the rental portal usage. Map the top three daily frictions with specific dollar attribution. Output: integration scope + flagship agent selection.
API credentials provisioned in the Anthropic workspace. Sandbox integration established with CCC Secure Share or Mitchell Connect. Flagship agent Routine configured, prompt-tuned, and tested against 10–15 historical workfiles supplied by the shop. Output: working agent running against test data.
Trigger activated against live production data. Estimator trained on the session URL approval workflow (15-minute walkthrough). Monitoring dashboard instrumented. Daily check-ins for the first five business days. Output: agent producing live recommendations, estimator approving or declining each one.
First monthly ROI report delivered: lines captured, hours saved, dollars recovered. Debrief call to identify next pain point for Stack or Operator tier expansion. Shop decides to stay on Reflex, upgrade to Stack, or jump directly to Operator. Output: the shop has proof in hand and a path to more.
Clarity on what this engagement does not do is as important as what it does. Boundaries protect the shop from over-promised software and protect Intersection Strategies from scope creep that dilutes outcomes.
We integrate with CCC ONE, Mitchell RepairCenter, and Audatex Qapter. We do not replace them. The estimating database remains the shop's system of record. Agents operate around the SMS, not in place of it.
We do not tell the technician how to fix the vehicle. Agents are administrative — they handle documentation, supplementation, and portal friction. Diagnostic and structural decisions remain with the licensed technician and I-CAR-certified repair planner.
Agent 04 generates documentation that supports mechanics lien defense. Agent 05 generates EV compliance logs. Neither is a substitute for retained counsel. When litigation arrives, the documentation helps — the attorney still prosecutes.
The mechanical path from interest to live agent runs through five checkpoints. All five happen inside the first ten business days. Nothing requires the shop to hire, provision, or deploy anything internally.
A focused conversation to confirm the SMS, the DRP mix, and the immediate pain. No pitch deck. The flagship agent is selected before the call ends.
Standard master services agreement plus the data-sharing exhibit. Intersection Strategies operates under NY DFS Part 500 TPSP compliance — the paperwork protects the shop's handling of nonpublic information.
Shop provisions read/write access to CCC Secure Share, Mitchell Connect, or Audatex APIGateway as applicable. ARMS/HIRS portal credentials shared via password manager. Typically one to three business days.
A two-hour technical window with the shop's IT contact (or CCC/Mitchell support rep). All integration work happens cloud-to-cloud — nothing installed on shop hardware.
Agent fires on the first qualifying production trigger. First session URL lands in the estimator's Slack or email. The capture begins — and the monthly ROI report starts accumulating from day one.
The agent stack is already configured. The integrations are already tested. The ROI math is already documented. The only missing piece is the first production trigger — and that is a fifteen-minute conversation away.
Schedule Discovery Call →