[ AFTERSALES & DEALER OPERATIONS ]LIVE
Hub Customer Service for Automotive Dealerships
Service that calls the customer before the customer calls the dealer. Hub identifies vehicles heading for a fault from connected-vehicle diagnostics and service history, works out which parts the repair needs, checks them against workshop inventory, and briefs the service adviser before the outreach call is made.
Monitored assets
| VIN | Customer | Model | DTC | Risk | Trend |
|---|---|---|---|---|---|
| …7HK2210 | R. Mehta | 2023 SUV 2.0T | P0420 · catalyst | Call today | |
| …3QL0087 | A. Fischer | 2022 EV LR | B1A4F · 12V battery | This week | |
| …9PX4412 | L. Chen | 2024 Hatch 1.5 | C1233 · ABS sensor | This week | |
| …2ZD1190 | S. Okoro | 2021 Saloon 2.0D | P2002 · DPF | Monitor | |
| …5MN7734 | J. Park | 2023 SUV 2.0T | P0300 · misfire | Monitor | |
| …8RT0921 | D. Novak | 2022 EV LR | U0111 · BMS comms | Resolved |
…7HK2210 · adviser brief
- 01P0420 logged 9× in 14 days; efficiency trend crossing threshold.
- 02Repair per TSB-23-114: catalyst + upstream O2 sensor. 2.4h labour.
- 03Parts: CAT-2210 (2 in stock), O2S-118 (arrives Thu) → book Fri.
- 04Customer: prefers mornings, last visit 5★, declined tyres in March.
// Problem
The Problem
Dealer aftersales is reactive by construction: the customer notices something wrong, calls, waits for an appointment, arrives, and only then does the workshop discover that the part is a three-day order. Connected vehicles have been emitting diagnostic trouble codes the whole time, and the service history has been showing the pattern, but neither is joined to the CRM record or to the parts inventory. The result is a repair that takes a week and a customer who experiences the dealer as slow rather than as attentive.
- Diagnostic trouble codes from connected vehicles are collected but never joined to the customer record.
- The parts a repair will need are established after the vehicle is already on the ramp.
- Repair time is set by parts availability nobody checked, so the courtesy-car queue absorbs the failure.
- Outreach, when it happens, is generic — the adviser has no history-specific talking points.
// Overview
Hub joins three sources that dealers already own but rarely combine: historical service data, the CRM record, and OEM telemetry from the connected vehicle including diagnostic trouble codes. From that it identifies customers whose vehicles are showing a developing issue and constructs the repair before the conversation: DTC repair manuals are searched for the procedure, the required parts are extracted, and those parts are compared against workshop inventory at the relevant location. Only then is the customer contacted — with the adviser holding a profile built from prior interaction notes and data-driven talking points, and with the parts already confirmed in stock, so the booked repair completes in a single visit.
// AI System
Why AI
The parts step is the one that cannot be done with rules. A diagnostic code maps to a repair procedure that lives as prose and diagrams inside an OEM manual, and the parts list has to be read out of that procedure rather than looked up in a table. Equally, interaction notes are free text written by whoever took the last call, and turning a stack of them into a usable customer profile is a summarisation problem. Everything downstream — inventory check, scheduling, outreach — is ordinary software; it just could not start until those two reading problems were solved.
// Specs
Specifications
- SIGNALS
- Connected-vehicle DTCs, service history, CRM record
- PARTS
- Extracted from DTC repair manuals, checked against workshop inventory
- OUTREACH
- Adviser-led, with generated profile and talking points
- SCOPE
- Proactive detection through to booked, parts-confirmed repair
- AUTONOMY
- Detection and preparation automated; contact stays human
// Features
Features
- 01Customers with developing vehicle issues identified from telemetry, history and CRM together.
- 02Repair procedures read out of DTC manuals and turned into a concrete parts list.
- 03Required parts checked against workshop inventory before the customer is called in.
- 04Customer profiles assembled from prior interaction notes for personalised outreach.
- 05Data-driven talking points supplied to the service adviser at the point of contact.
- 06Customised communications sent from the dealer, under adviser control.
// Architecture
Architecture
AFTERSALES FLOW
Runtime · one item, left to right
- 01Vehicle Telemetry + Service History
- 02Issue Detection
- 03Repair & Parts ResolutionDTC CodesRepair ManualsParts InventoryCRM & Interaction Notes
- 04Inventory Confirmation
- 05Adviser Outreach
- 06Single-Visit Repair
dashed = the inference step, where the system exercises judgment
System stack
Data in · decisions out
01
Sources
Vehicle, workshop, customer
02
Ingestion
Join VIN to everything
03
Ontology
The vehicle as an object with history
04AI
Intelligence
Predict, resolve, brief
05Human
Human control
Adviser owns the relationship
06
Actions
Written back
Observability
Every model call traced; evals run on real cases, not anecdotes.
Governance
Entitlements enforced at retrieval; rules versioned by the organisation.
Write-back
Systems of record are written only through the approval gate.
No customer is called until the parts for the identified repair are confirmed on site.
// Impact
Impact
- Before the call
- Parts availability confirmeddesign intent
- One visit
- Target repair path, against a return tripindicative target
Interested in dealership customer service?
Let's look at what your connected fleet is already telling you.
Get in touch