Ducker Carlisle on streamlining automotive enterprises using Generative AI

Ducker Carlisle is a global consulting and M&A advisory company with an unrivaled continuum of insights, benchmarking, and strategy solutions.

The company optimizes business performance and accelerate growth for many of the world’s largest companies and private equity firms.

Automotive Industries (AI) sat down with the firm’s Fabien Cros, Chief Data & AI Officer at Ducker Carlisle to throw some light on their solutions aimed at streamlining and optimizing the 21^st century automotive enterprises using Generative AI and more.

Cros served as Data & AI Country Lead for Manufacturing at Google before joining Ducker Carlisle.  

AI: Early adopters reported a US$36 million uplift from improved RFP (request for proposal) wins. How do BidMaster or QuoteMaster change the sales cycle to produce that result?

Cros: At Google, I saw firsthand how AI can fundamentally shift a professional’s focus from mundane tasks to high-value strategic work. That’s the core principle behind BidMaster and QuoteMaster.

These tools don’t just automate proposal writing; they act as a strategic co-pilot for the sales team. They handle the tedious, time-consuming tasks: pulling client information, researching product specifications, and drafting the initial proposal.

This strategic shift is where the value lies. Instead of spending 99% of their time on a document, our clients’ sales teams now reallocate that energy to understanding a client’s genuine needs, conducting in-depth exploration sessions, and, most importantly, building human relationships that ultimately close deals.

This isn’t just about speed; it’s about enabling your team to be relationship-driven instead of process-driven. The $36 million uplift for one client wasn’t just a byproduct of faster proposals; it was the direct result of their team having the time to focus on influencing the deal at every stage.

AI: Many automotive firms run on fragmented legacy systems. How does SparkWise Librarian extract and normalize cross-system data reliably, and what are the most common surprises you uncover during integration?

Cros: This is the precise problem the SparkWise Librarian was designed to solve. As a data and AI professional, I know that data is the lifeblood of any modern operation, but it’s often trapped in silos.

The Librarian is powered by advanced AI solutions that are exceptionally good at treating and understanding both data and metadata.

It doesn’t just extract information. It cleans, sorts, and harmonizes it, creating a unified, reliable source of truth from disparate legacy systems.

The most common and, frankly, mind-blowing discovery we make during integration is that none of a client’s systems were 100% accurate.

They are only partially true. In the automotive and industrial world, where every approximation can have a ripple effect down the supply chain, it’s a revelation for our clients to realize how many errors exist in the data they’ve based their analyses on. SparkWise brings that data into focus, allowing for truly informed decisions for the first time.

AI: PriceGenius promises SKU-level pricing for spare parts. What data sources and models power its pricing recommendations, and how do you prevent model drift when market conditions change rapidly?

Cros: PriceGenius is highly customizable, and its power comes from its ability to ingest and synthesize a wide array of data sources that our clients choose to use, including CAD files, PIM systems, external benchmarks, and internal sales data.

Our approach to pricing is a bit different. We rely on a robust deterministic model for the core price calculation. This model is based on established rules and multiple scenarios, ensuring a stable foundation.

For market fluctuations, we leverage Generative AI to collect real-time data and add context, helping the model pick the most appropriate price from its deterministic calculations.

In the automotive parts sector, where prices generally change annually, we have plenty of time to fine-tune our models.

For e-commerce, where prices are more dynamic, our deterministic approach prevents model drift.

We also use a continuous feedback loop and, crucially, a “human in the loop” system to validate the AI’s recommendations. This ensures that every output is checked for accuracy and relevance before it’s implemented.

AI: You offer deployments in as little as eight weeks. How do you balance that cadence with clients’ needs for customization, internal approvals, and industry-specific regulatory or contractual constraints?

Cros: The eight-week timeframe is the development and customization period on our end. My experience at Google taught me the power of a lean, user-centric product methodology.

We build with the end-user in mind from day one, which streamlines the development process significantly. Our pre-packaged solutions have a core backbone of code, testing sets, and user prompts that we’ve already developed and tested with similar companies.

This allows us to deploy rapidly while still providing the necessary customization.

If a client’s internal approval processes require more time, we simply adjust our pace to match theirs. We are flexible partners, but we have proven that the core solution can be built and deployed in eight weeks for clients who are ready to move quickly.

AI: Hosting solutions in a client’s cloud mitigates some cyber risk. What threat scenarios worry you most for automotive customers using Generative AI, and what controls and governance does SparkWise enforce by default?

Cros: Cyber security is non-negotiable, and we work with a very simple principle: our clients’ data is their data. We offer to host the solution directly within their secure IT environment, leveraging their established cybersecurity and IT requirements.

Alternatively, we can use a VPN for the project, ensuring a secure connection to our systems.

Our confidence comes from the robust security of hyperscalers like Google Cloud and Azure, which we use as a foundation. Generative AI, while powerful, can pose risks if not properly managed, primarily around the handling of proprietary data during model training.

Our governance protocols ensure that no client data is used to train a general model and that it remains entirely within their secure environment, whether that’s on our side of the VPN or within their own cloud.

AI: For private equity portfolio companies and OEMs alike, what concrete KPIs and timeframes should leaders expect to measure to validate the top-line and bottom-line claims you’re making?

Cros: We are completely transparent about our value proposition and timeline. Leaders should expect to measure two key phases.

The first is the initial eight-week deployment, after which we guarantee our clients will see approximately 80% of the solution’s full value. This is because we build solutions that are user-centric and address immediate pain points, leading to instant adoption and value creation.

The second phase, which spans six to ten months, is for fine-tuning and optimization. During this period, we work together to tweak the solution to capture the remaining 20% of value.

This methodical approach ensures that our clients don’t have to wait years to see a return on their investment.

AI: Automation often triggers organizational friction. How does Ducker Carlisle approach change management and upskilling when SmartFlow or ContractNavigator replaces manual roles or processes?

Cros: I’ve found that organizational friction isn’t just a byproduct of change; it’s often a symptom of misaligned incentives. You cannot expect people to fully embrace 21st-century tools if you’re still using 19th-century compensation and incentive structures.

A major part of our project is to help clients rethink their organizational models and reward systems.

Furthermore, we utilize the same user-centric product methodology I learned at Google. We don’t just build solutions and then hand them off.

We involve the end-users from the very beginning, building and iterating with them. This collaborative approach makes onboarding seamless because they’ve been part of the journey. The AI is built to solve their pain points, making adoption intuitive rather than a forced transition.

AI The market now has cloud hyperscalers, specialist AI vendors, and in-house teams. What competitive advantages does SparkWise bring (tech, IP, industry domain, delivery model) that make these eight offerings a better fit for automotive clients?

Cros: We are a unique hybrid of builder and expert. Our competitive advantage lies in our ability to combine deep industry and automotive expertise, courtesy of Ducker Carlisle, with our unparalleled tech, product, and AI know-how from SparkWise Solutions.

We are not just a tool provider; we are problem-solvers who use all the available technologies on the market to create a personalized, pain-point-specific solution.

Our eight pre-packaged solutions are built on a solid foundation of existing IP, including a backbone of code, tested sets, and user prompts that have already been developed with similar companies.

This unique model allows our clients to benefit from a proven, pre-vetted solution that is still fully customizable to their specific needs.

If you would like to find out more visit    https://www.duckercarlisle.com/industries/automotive-transportation/