The snapshot
1
Strawberry, a leading Nordic hospitality group, faced severe information overload as vital operational knowledge became trapped across fragmented legacy systems and scattered documents.
2
To resolve this, Devoteam and Google Cloud engineered Scout AI, an enterprise-grade Agentic AI assistant utilising a secure Retrieval-Augmented Generation (RAG) architecture.
3
The solution was seamlessly integrated into existing communication platforms like Google Chat and Slack, delivering instant, verified answers without requiring disruptive data migrations.
About the Customer: A Nordic Hospitality Innovator
Strawberry (formerly Nordic Choice Hotels) operates more than 230 hotels across the Nordics and Baltics. With roughly 18,000 employees, the company’s mission is to deliver “a better world through travel.” For over 25 years, they have set the standard for guest experiences. You will recognise them for their tech-forward mindset, treating digital transformation as a core pillar of their frontline service.
The Challenge: Conquering Information Overload
When you continually upgrade systems across hundreds of hotels, you inadvertently create an ocean of new procedures. Strawberry’s vital knowledge became trapped across scattered PDFs, varied intranets, and complex software systems.
Historically, employees relied on manual searches and internal phone calls to track down basic policies. This outdated approach was slow and error-prone. It distracted the frontline staff from what mattered most: the guests. New joiners faced an overwhelming learning curve, and information overload became a serious bottleneck to smooth, efficient operations.
The Goal: Instant Knowledge at Your Fingertips
Strawberry wanted to eliminate the endless searching. The goal was to build a centralised, highly accessible platform where the staff could instantly find accurate, verified company knowledge.
They envisioned a digital colleague. They needed an intelligent assistant that could deliver real-time, multilingual answers about company culture, FAQs, and hotel operations, all without forcing employees to learn a complicated new system.
As a company, we have a strong focus on sustainability and digital innovation—for our customers as well as our employees. We want to give employees sustainable technology to help with their day-to-day functions while providing guests the best experience possible.
Kari Anna Fiskvik,
Chief Digital & Technology Officer at Strawberry
The Solution: Building Scout AI with Google Cloud
Devoteam proposed Scout AI, an enterprise-grade assistant built entirely on Google Cloud. We utilised Gemini to power the AI’s reasoning and LangChain to create a Retrieval-Augmented Generation (RAG) architecture. This ensured the AI only retrieved information from Strawberry’s verified data, preventing hallucinations.
Why this solution? Strawberry needed to meet users exactly where they already worked. Scout seamlessly integrated into Google Chat and Slack via function calling. The team did not have to change platforms or undergo disruptive data migrations.
What made this possible? The deep partnership. Devoteam previously helped Strawberry recover from a major ransomware attack by converting 3,000 computers to ChromeOS in less than a week. Additionally, Devoteam built a custom AppSheet environment for security audits. Devoteam understood Strawberry’s Google Workspace ecosystem, allowing the total solution to remain flexible and highly secure.
Early signs of success were immediate. By giving Scout a deliberately “sassy” and friendly personality, employee adoption skyrocketed. For example, when an employee like Adam asked a complex, multilingual policy question directly in Google Chat, Scout instantly pulled the verified answer from internal APIs.
The Methodology: A Collaborative Path to Agentic AI
Devoteam and Google Cloud drove this implementation using a highly collaborative, agile framework.
First, the team ran a discovery phase to map data silos and identify critical bottlenecks. Next, they launched a Minimum Viable Product (MVP) within just two weeks to test accuracy and gather real-world feedback on the user experience.
Finally, Devoteam deployed the Agentic AI framework using Vertex AI Agent Builder. The team tuned individual “Sister Agents,” such as the guest-facing Lykke, to match the distinct persona of specific hotels. By expanding the AI engine to guest-facing agents like Lykke, visitors receive 24/7 support. For more on optimising these interactions, read our expert view on AI agents for customer service. Experts also established a continuous telemetry loop to flag missing data. If multiple users searched for a specific topic, the system immediately alerted the marketing team to update the documentation, continuously closing content gaps.
Hear the success story straight from the source
We handed the microphone over to our client at Strawberry! Watch this quick recap as Evelyn Piros-Read, AI Product Manager at Strawberry, shares their experience.
The Result: Reclaiming 36,000 Hours for People
25,000 hours saved for employees
11,000 hours returned to guests
98% positive guest sentiment
By transforming scattered documents into a unified knowledge brain, Scout AI fundamentally changed how the teams operate. They shifted their business from manual searching to proactive, AI-first hospitality.
The visible consequences were immediate and highly measurable:
- 25,000 hours saved for employees: Staff now resolve queries in seconds instead of minutes, minimising IT support tickets and drastically reducing administrative downtime.
- 11,000 hours returned to guests: By expanding the AI engine to guest-facing agents, visitors now receive instant, 24/7 expert support without queuing at the front desk.
- 98% positive guest sentiment: The accuracy of the RAG architecture ensured high-quality interactions, keeping satisfaction scores exceptionally high.
Your Success Starts Here
Partner with Devoteam for award-winning tech expertise, agile execution, and a culture of continuous learning.