Each colour maps a published value to the exact source passage used to support it. Only bounded excerpts are public; administrators can inspect the complete captured source.
Title
ICA Gruppen Cuts AI Proof-of-Concept Time from 18 Months to 10 Weeks on BigQuery
…l business Improves ML model training and deployment speed while lowering costs ICA modernized its data architecture on Google Cloud to break down silos and accelerate AI innovation from 18 months to 10 weeks. Scaling a century of service through a unified data vision ICA has been a corne…
Description
Swedish retailer ICA Gruppen architected its Business Intelligence Cloud Analytics (BICA) platform on Google Cloud, centered on BigQuery, and migrated its machine learning workloads, including its personalized recommendation engine, to the Gemini platform. The team also launched Shop with AI, a Gemini-based shopping assistant that simplifies meal planning. A recent proof of concept with ICA Bank was completed in two and a half months, a process that would have taken approximately 18 months previously.
…servers. This architectural agility has transformed ICA’s ability to prototype. A recent proof of concept with ICA Bank was completed in just two and a half months—a process that would’ve taken approximately 18 months previously. This development speed increase is possible because the data and technical capa…
Company
ICA Gruppen
…where data-driven AI innovation makes every day easier for millions of people. ICA Gruppen is a leading Swedish retailer, with a portfolio covering grocery, pharmacy and banking. Serving more than 5M people daily, its mission is to make every day a little ea…
Country
Sweden
…ission is to make every day a little easier for its customers. Industry: Retail Location: Sweden Products: BigQuery , Bigtable , Gemini , Gemini platform , Looker , Spanner men…
Industry
Retail
…ople daily, its mission is to make every day a little easier for its customers. Industry: Retail Location: Sweden Products: BigQuery , Bigtable , Gemini , Gemini platform , Loo…
Problem
ICA's fragmented technical environment was holding it back: data was siloed across legacy platforms and SQL server environments, creating high maintenance costs and preventing teams from sharing insights, while innovation stalled as data scientists struggled to push ML models into production.
…reviously, however, ICA’s fragmented technical environment was holding it back. Data was siloed across legacy platforms and SQL server environments, creating high maintenance costs and preventing teams from sharing insights, while innovation stalled as data scientists struggled to push ML models into production. To improve operational efficiency and provide a better customer experience, ICA…
Solution
ICA architected the Business Intelligence Cloud Analytics (BICA) platform on Google Cloud, centered on BigQuery, moving away from traditional ETL tools toward an engineering-led approach with its own code-based frameworks to automate ingestion, security and connectivity. It streams high-volume store data through Kafka into BigQuery, uses a custom column-level encryption service with decryption views so unencrypted data is never stored at rest in BigQuery, and uses Spanner and Bigtable to power low-latency operational APIs. ICA has transitioned to Looker as its semantic layer. It migrated its machine learning workloads, including its personalized recommendation engine, to the Gemini platform, and used the data foundation to launch Shop with AI, a shopping assistant that uses Gemini to let users chat to simplify meal planning, combining ICA's recipe library with real-time customer insights.
…tor data with BigQuery To resolve data bottlenecks and pursue its AI ambitions, ICA architected the Business Intelligence Cloud Analytics (BICA) platform on Google Cloud , centered on BigQuery . This shift represented a move away from traditional, restrictive Extract, Tra…
…ogle Cloud. To serve this security infrastructure without introducing friction, ICA uses Spanner and Bigtable to power low-latency operational APIs. While analytical workloads run on BigQuery through a 15-minute micro-batching w…
…ion keys, ensuring instantaneous access without impacting platform performance. To ensure everyone operates from a single source of truth, ICA has transitioned to Looker as its semantic layer. Looker provides a central entry point to the company data, where business metri…
Business value
Migrating its machine learning workloads to the Gemini platform improved training and deployment speeds while lowering operational costs. A recent proof of concept with ICA Bank was completed in just two and a half months, a process that would have taken approximately 18 months previously, and the team says it can move from an idea to a real-world pilot in weeks rather than months.
…workloads—including its personalized recommendation engine—to Gemini platform , improving training and deployment speeds while lowering operational costs. Gemini platform provides a managed environment that handles the underlying inf…
…shared infrastructure, the threshold for a proof of concept is incredibly low. We can move from an idea to a real-world pilot in weeks rather than months. Heaven Bereket Business Translator, Data Science and AI, ICA The ICA team also…
AI capabilities
Generative AI, Conversational AI
…and AI, ICA The ICA team also used this data foundation to launch Shop with AI, a shopping assistant that uses Gemini to allow users to interact through chat to simplify meal planning. The agent combines ICA's recipe library with real-time customer insights to ge…
Technology
BigQuery, Bigtable, Gemini, Gemini platform, Looker, Spanner
…every day a little easier for its customers. Industry: Retail Location: Sweden Products: BigQuery , Bigtable , Gemini , Gemini platform , Looker , Spanner menu Overview Solutions Products Pricing Resources Docs Support Contact us …
Headline outcome
…servers. This architectural agility has transformed ICA’s ability to prototype. A recent proof of concept with ICA Bank was completed in just two and a half months—a process that would’ve taken approximately 18 months previously. This development speed increase is possible because the data and technical capa…
Use case type
Personalisation
…bility to innovate with AI. The company migrated its machine learning workloads—including its personalized recommendation engine—to Gemini platform , improving training and deployment speeds while lowering op…
