Automotive Software Innovations

Last weeks All Hands was another great reminder of how fast our teams are turning AI from an experiment into something that’s really shaping how we work every day. A few examples from just the past days: 🔹 Competitor Co-Pilot: Helps our sales and internal teams analyze competitor pricing, terms, and workflows across several legacy players, with full document traceability. 🔹 Customer Insights Model: Automatically summarizes interview data to benchmark service quality (like returns, document handling, and complaints) and retrains itself with new feedback. 🔹 Image Recognition in Vehicle Document Scanning: Now used to make dealership document processes faster and more accurate. 🔹 Trade-In Co-Pilot Agent: Assesses vehicles by their unique ID, showing which features add or reduce value and highlighting common issues for similar cars. Great to see how much progress can happen when everyone contributes their own ideas and experiments with new tools.

5 Agentic AI applications I would prioritize if I was Head of Quality at an Automotive OEM today. 1) Field Signal Detection from Public Data: Monitor social media data, reviews and dealer complaints for emerging failure patterns. Surface the product issues in week 3 to take immediate action to reduce future warranty costs. 2) Warranty Pattern Clustering from Service Records: Technician repair notes are written in shorthand, abbreviations, and inconsistent language across thousands of dealers. An agentic AI system reads every repair note, normalizes the language, clusters similar complaints, and surfaces failure patterns that manual review consistently misses. The pattern hiding in 50,000 repair notes is not visible to any human analyst. It is visible to an agent running continuously. 3) Supplier Risk Monitoring from External Signals: Financial distress, senior departures, declining shipment volumes, regulatory actions. These signals appear in public data months before a supplier quality failure reaches your incoming inspection. An agent monitoring your top 50 suppliers across these signal sources continuously gives your procurement team early warning of supplier failure. 4) Dealer Service Knowledge Propagation: When one service center solves a recurring fault, that fix stays in that workshop. Hundreds of other centers struggle with the same fault next month. An agent reads repair notes across the entire dealer network, identifies when one center has solved problem others are still struggling with, and propagates that fix automatically. Faster repairs. Lower warranty labor costs. Better customer experience. 5) Telematics Data for Quality Intelligence: Vehicle telematics generates enormous volumes of real-time data on battery performance, motor behavior, charging patterns, and component stress across the entire fleet. An agentic AI system continuously monitors this data to identify components showing early signs of degradation before owners report symptoms and before warranty claims are filed. The failure pattern exists in the telematics stream weeks before it surfaces anywhere else. Which of these would you prioritize first? #AutomotiveAI #AgenticAI #QualityManagement #WarrantyAnalytics #ManufacturingAI

The biggest AI opportunity in automotive retail is not sales. It is after-sales. Why? Because after-sales is where profitability, retention, and customer trust actually compound. Most dealer groups still use AI at the top of the funnel: chatbots, lead qualification, WhatsApp automation. Useful, yes. Transformational, not yet. The real value starts when AI is embedded in the service lane: - Predicting the right moment to contact each customer. - Automatically booking the next service visit. - Reducing no-shows. - Prioritising high-value jobs. Recommending the next best action for each vehicle and owner. This is where the economics change. A small improvement in workshop occupancy, repair order conversion, or upsell rate can have a much larger impact than many incremental sales-focused tools. And unlike many brand campaigns, these gains are measurable quickly. The challenge is not technology. It is operational design. Most dealerships do not fail because the model is weak. They fail because: - The data is fragmented. - The process is not redesigned. - The team is not trained. - The KPI is not owned by management. If AI is not connected to the DMS, the service agenda and the advisor workflow, it becomes an “extra tool” instead of a productivity engine. For OEMs, NSCs and dealer groups, the right question is not “Should we use AI in after-sales?” It is: Which service process should AI improve first to create margin fast? In my view, the answer is usually: - appointment booking, - recall and maintenance reminders, - workshop scheduling. That is where AI becomes real. Hartmut Wagner Pedro Sala Peter Petrovski Jens Monsees Ricardo Oliveira Javier Garcia Christian Zamet MERCEDES MONDELO ALONSO Elvira Llorens Ian Plummer David Ortega Joan Miquel Malagelada Juan Montesinos Hernando José Antonio Soria Rodríguez Pierre Boutin Laura Ros Verhoeven José Miguel Aparicio Iñaki Nieto Liliana Gachancipa Correa Paul Bennett Mayte Mercader Rosco

How to pilot AI in your business without risking your reputation or budget (based on 10+ successful implementations) Everyone's rushing to implement AI. 80% will fail in the first 3 months. Here's how to be in the successful 20%. The 3-Phase Pilot Framework: 1. The Safety Zone Phase → Internal processes only → No customer interaction → Limited data access → Small team exposure Real Case: A Mercedes dealership started with internal document processing: → 100 documents daily → 2 departments → 5 team members Result: 90% accuracy before any customer exposure 2. The Controlled Testing Phase → 20% of customer interactions → Parallel human oversight → Real-time monitoring → Immediate fallback options 3. The Safe Scaling Strategy → 0% customer exposure week 1-2 → 20% customer exposure week 3-4 → 50% customer exposure week 5-6 → Full implementation week 7-8 Real Case Results: Real estate AI implementation: → Started with 50 properties → Tested with 3 agents → Achieved 97% accuracy → Scaled to full operation in 60 days The Budget Protection Strategy: Phase 1: Discovery → Fixed cost: $5-10K → Clear go/no-go criteria Phase 2: Pilot → Controlled budget → Weekly review points Phase 3: Implementation → Success-based scaling → Predictable monthly costs Remember: The goal isn't to be first with AI. The goal is to be right with AI. Want to explore a risk-free AI pilot for your business? DM me for a 30-minute consultation where I'll help you design a safe implementation strategy based on your specific needs.

Most dealerships don’t lose leads at the top. They lose them between steps. I saw this clearly while working on automotive workflows at EaseZen. Leads were coming in. Sales teams were busy. But follow-ups slowed down, bookings slipped, and CRM data was always behind. The problem wasn’t effort. It was manual execution. So instead of adding more tools, we rebuilt the funnel to run itself: Lead capture → follow-up → test drive booking → CRM update That’s the funnel you see in the visual 👇 ① Lead Capture Every website, ad, or chat lead flows straight into Frappe CRM. An AI agent qualifies interest instantly. No waiting. No missed inquiries. ② AI Follow-Up AI scores intent and responds automatically via email, WhatsApp, or SMS — at the right moment. Hot leads move fast. Cold leads stay warm. ③ Test Drive Booking Once intent is clear, AI schedules the test drive. No chasing. No back-and-forth. ④ CRM Update & Close Every action syncs back into the CRM automatically. Clean stages. Accurate dashboards. Behind the scenes, this runs on a simple stack: Frappe CRM + n8n automation + agentic AI agents Each step is owned by AI. Humans step in only for judgment. That’s the shift most dealerships miss. If your funnel feels busy but stuck, it’s not a lead problem, it’s an execution problem.

I was asked a smart question at an NCM 20 Group this week. “If you were a Dealer Principal again, where would you attack with AI first?” No hesitation. Existing customers. Service drive. Not marketing. Not shiny new tools. Not more leads. Here’s why. Service hits the bottom line fast. It’s already profitable. And it touches your most valuable customers more often than sales ever will. AI applied here does three things immediately: Identifies your most valuable customers Not by gut feel. By actual behavior, spend, loyalty, and retention patterns. Protects fixed ops revenue Who’s drifting. Who’s skipping visits. Who’s one bad experience away from disappearing. Trickle-feeds sales without forcing it Trade signals. Equity timing. Mileage and ownership triggers. All discovered inside the service lane, not the showroom. This is the mistake I see dealers make with AI. They start where it feels exciting. Instead of where it compounds. Service is your data goldmine. Your cash engine. Your safest place to learn how AI actually works inside your business. If I were running a store again, I’d earn my AI ROI there first. Then scale outward. Curious where you’d start if you had to pick one department.