Skip to main content
AmigoFish

How Carwow shifted paperwork from manual to automatic to save £2M a year

From Google Cloud · · 2 min

The player loads from YouTube only when you click. It uses YouTube's privacy-enhanced mode.

If the video does not play, the owner may have turned off embedding. Watch it on YouTube instead.

Watch on YouTube

Not rated yet

Description

*Featured in this video:* Becky Allsop, Director of Analytics and Data Science at Carwow

*Executive summary:* Carwow, one of Europe's largest online car-changing destinations, partnered with Google Cloud to shift its AI strategy into high gear. Using Gemini for business, Carwow automated document verification, automatically identifying and extracting key information from complex vehicle records while protecting sensitive data — an early win that delivered £2 million in annual savings. Building on this momentum, Carwow is now using BigQuery, Looker, Knowledge Catalog, and Gemini platform to establish a strong AI foundation to power its next phase of transformation, enabling a single source of truth across the business and cutting model deployment cycles from months to weeks without engineering reliance. Looking forward, Gemini platform will also provide a launchpad for more advanced use cases: Carwow is already actively exploring generative models like Chirp and Veo to repurpose and localize high-performing content from its popular YouTube channel.

*Challenge:* Carwow built its brand by replacing intimidating dealership visits with transparent digital experiences. However, sustaining this advantage means continuing to extend its trailblazing portfolio to reach drivers everywhere. Carwow wanted to rebuild with AI at its core to continue providing the choice, data, and opinion people need to feel confident in every car change.

*Solution:* Carwow's first AI win delivered £2 million in annual savings, using Gemini for business to identify and extract critical verification details from complex documents — like handwritten service receipts and official vehicle logs — all while automatically redacting sensitive customer data. Building on this success, Carwow rebuilt its MLOps pipeline on Gemini platform, providing a consistent framework for building, testing, and monitoring models without ML engineering reliance. Carwow is also migrating to BigQuery and deploying Looker to enable governed, self-service analytics across the business, using Looker’s semantic layer and Knowledge Catalog to standardize business context and automate data governance and quality tracking. Gemini platform provides access to the latest AI capabilities, including generative models like Veo and Chirp, serving as a launch pad for more advanced use cases to scale brand reach across global audiences. For instance, Carwow is actively exploring how cutting-edge Google generation models like Chirp and Veo can be used to repurpose high-performing content from its popular YouTube channel hosted by automotive journalist turned car influencer — Mat Watson — which boasts over 11 million subscribers.

*Results:* With Google Cloud, Carwow has the foundation it needs to build the car-changing destination of choice for the AI era. Now, data scientists can independently maintain and manage the entire model product lifecycle, cutting deployment cycles from months to weeks. At the same time, establishing a single source of truth allows Carwow to understand, govern, and manage all its data in one central place, delivering the trusted foundation required to power its long-term AI vision.

*Key takeaways and highlights from our interview with Becky Allsop, Director of Analytics and Data Science at Carwow:*

→ “We’ve used Gemini to automate [document verification], even the handwritten notes. It verifies and identifies PII and can redact that so we're not exposing personal data where we shouldn't be. It's been a massive win both from an internal perspective for colleagues but also from a customer perspective. It makes that process so much quicker, and we’ve seen huge savings. It’s up to £2 million a year in savings.”

→ “We’ve recently done a project to completely rebuild our MLOps platform with Gemini. We're now at a point where we can literally move from proof of concept to live within a matter of weeks. The framework we've built has actually meant we’re not relying on ML…

Back to Google Cloud · Report a problem