95% accuracy in automated bill data extraction thanks to AI
+30% conversion rate: bill uploaders versus non-uploaders
+60% conversion rate: chatbot users versus non-users
Up to 90% reduction in manual process time
AI democratized: non-technical teams deploy workflows autonomously
Facile.it uses Google Cloud’s AI technologies to automate bill and doc processing at 95% accuracy, boost conversions, and democratize AI across the units.
Facile.it uses Google Cloud’s AI technologies to automate bill and doc processing at 95% accuracy, boost conversions, and democratize AI across the units.
From one of Italy's leading comparison platforms to an AI-first company
When one of the main Italian comparators Facile.it introduced the ability for users to upload their bills and contracts directly on the platform, letting AI extract all relevant fields automatically, conversion rates jumped +30% compared to users who didn't use the feature. That result captured the core challenge the company was trying to solve: users comparing insurance, energy, and financial products were losing patience with long forms, while the platform needed more data, not less, to generate accurate recommendations. AI was the only way to resolve that tension.
The extraction pipeline runs on Document AI, Google Cloud's document understanding service, which processes uploaded bills and contracts at 95% accuracy, according to Facile.it. Data flows into Google Cloud's BigQuery, which serves as the company's central analytical warehouse.This enables the platform to continuously refine its recommendation models based on real user data.
"We reached a point where we understood that simplicity alone wasn't enough" says Luca Boeri, VP of Technology at Facile.it "We needed to eliminate the trade-off between asking users for more data and more time and making their experience easier.
We were asking users to spend more time tracking down all the right fields, which in financial contracts are many, in exchange for greater savings. With AI, that trade-off is no longer necessary.
Luca Boeri
VP of Technology, Facile.it
We were asking them to spend more time tracking down all the right fields, which in financial contracts are many, in exchange for greater savings. With these optimisations, that trade-off is no longer necessary, and we have more data to give them a better service right away, and in the future too."
The same logic applied internally. AI capabilities existed but were concentrated in technical teams, creating a bottleneck that slowed innovation across the business. Facile.it is one of Italy's largest online price comparison platforms, with over 5 million monthly users and more than €200 million in annual revenue. It had completed a full migration to Google Cloud in 2023. The infrastructure was in place. The next move was strategic: evolving from a cloud-native company into an AI-first one, embedding intelligence not as a feature but as a foundation across products, processes, and people.
AI agents: from smarter conversations to lifelong financial guidance
Facile.it deployed AI-powered conversational agents on its website, built with Google Cloud's Agent Development Kit and Gemini platform, as dynamic recommenders capable of understanding a user's profile and guiding them toward the right product in real time. The conversational agents are deployed on Gemini platform and handles Facile.it's scale of over five million monthly users without manual infrastructure management.
The real ambition is not to help users make a decision today. The next step is to use that knowledge proactively: to reach out when a better deal becomes available, before they even think to look.
Luca Boeri
VP of Technology, Facile.it
Users who engage with the chatbot and follow its recommendations convert at a rate 60% higher than those who navigate without assistance. But the conversion rate, as significant as it is, is not the end of the story.
"The real ambition is not to help users make a decision today" says Luca Boeri. "It's to stay with them over time, helping them starting from what we already know; their energy consumption, their insurance history, their financial profile. The next step is to use that knowledge proactively: to reach out when a better deal becomes available, before they even think to look."
That vision of Facile.it as a permanent financial companion rather than a one-time comparison tool, is what drives the company's AI investment beyond immediate conversion metrics. The platform is moving toward a model where AI is used to create a virtual consultant, capable of advising users by suggesting the best products or identifying when switching makes financial sense for them.
Enabling that future requires a foundation the company can trust. Facile.it uses Google Cloud's Model Armor to govern the behavior of AI agents exposed to end users, ensuring responses are safe, accurate, and appropriate for a regulated financial services context. Spending visibility across all AI services is managed in real time through the Google Cloud console, giving the team full control over what is running, who is using it, and what it costs. "Being inside Google Cloud gives us security not just over user data, but over our own operations” says Marco Saletta, Head of Data, Analytics and AI "That's non-negotiable in a sector like ours."
Democratizing AI: from individual productivity to organizational transformation
The third dimension of Facile.it's AI transformation is the most far-reaching. AI tools confined to specialist teams create a two-speed company. To avoid that, Facile.it made Gemini platform a standard tool in the hands of every developer, accessible directly from local environments, deployable to production through existing workflows. Non-technical teams followed: Through Google Workspace, Gemini and orchestration tools, non-technical teams in operations, marketing, and finance build and deploy complex workflow automations autonomously, without writing a single line of code. What previously required engineering support is now handled directly by the business units.
We think of this as three stages. First, AI increases individual productivity. Then it transforms how entire teams work and processes are redesigned with AI at the centre. Finally, it reshapes the organisation itself.
Marco Saletta
Head of Data, Analytics and AI, Facile.it
"We think of this as three stages" says Marco Saletta "First, AI increases individual productivity. Then it transforms how entire teams work and processes are redesigned with AI at the centre. Finally, it reshapes the organisation itself; roles change, structures change, and what was once impossible becomes routine. We are moving through those stages deliberately, and the pace is accelerating."
The financial services unit illustrates the second stage. Managing customer relationships across mortgages, loans, and investments had involved substantial backoffice effort.
Automations built with Gemini platform have cut manual processing time by up to 90%, turning hours into minutes without reducing the quality of the customer experience.
The third stage is taking shape in engineering, where AI agents now perform automated code review at a fraction of the previous time. But inside Facile.it, this is seen as an early signal of something larger: a future in which AI does not assist the engineering process but helps coordinate it.
Facile.it is a leading Italian online price comparison platform that allows users to compare prices for insurance products, utilities, and financial products. The company generated more than €200 million in revenue and has over 5 million unique monthly users.
Industry: Financial Services
Location: Italy
Products: DocumentAI , Gemini platform (Gemini 2.5 and 3 both Pro and Flash), BigQuery, ADK, GKE, Cloud Run, Google Workspace
