Meta Global AI Developer Hackathon. $1M in prizes.Get notified
LLAMA 3.1 405B
Brain4Data
80%
of customer service requests are handled by Llama-powered AI100%
multilingual — the agent converses in the customers’ language*All results are self-reported and not identifiably repeatable. Generally, expected individual results will differ.
CASE STUDY
Enhancing guest experiences with an AI customer service chatbot
At a glance
Industry
TechnologyUse case
Automating micro-hotel customer serviceGoal
Improve the guest experience and streamline operationsLlama version
Llama 3.1 405BDeployment
Oracle Cloud Infrastructure (OCI)THEIR STORY
Empowering businesses to put AI to work
Brain4Data specializes in helping businesses use advanced AI technology and retrieval augmented generation (RAG) to prepare data for generative AI applications. Their digital AI assistant, FRED, transforms distributed company information into high-value, AI-ready knowledge bases and generates precise, context-specific answers. Currently, FRED generates 40,000 responses a month for clients in 30 countries.Roatel Holding AG is the leading provider of smart micro-hotels in Germany. Self-service Roatel facilities provide travelers with convenient, affordable lodging with all the modern conveniences. Based in Germany, Roatel plans to establish a network of hotels across Europe.
THEIR GOAL
Automating guest service to unlock business growth
Brain4Data’s client, Roatel, operated self-service micro-hotels in Germany. Hotel guests relied on a remote customer service team to answer their questions, but as Roatel began to expand, the service team struggled to keep up with increased demand. Brain4Data set out to build a scalable AI solution that could automate customer service in multiple languages and provide accurate, fast responses 24/7. The goal was to maintain guest satisfaction and reduce reliance on live customer service agents.
THEIR SOLUTION
A Llama-powered customer-service chatbot
Brain4Data used its digital assistant platform FRED, Llama and OCI Generative AI to develop a chatbot for Roatel. The chatbot provides fast, reliable and up-to-date answers to complicated questions — no matter the time of day.Llama 3.1 405B is at the heart of the solution. Brain4Data chose the Llama model for its multilingual capabilities and strong performance in nuanced, semantic understanding — all delivered without fine-tuning. The team also knew the chatbot would benefit from an open-source model with broad community support.
The chatbot responds to guest requests in 11 languages at any time of day.
THEIR APPROACH
Streamlining development with OCI tools
The chatbot is built on OCI using OCI Generative AI and a RAG workflow. When a guest asks a question, the system gathers relevant context from Roatel’s document repository and injects it into a custom-engineered system prompt that is passed to the Llama model, which generates a response in the guest’s preferred language. Behind the scenes, Oracle’s Database 23ai Vector Search identifies the most relevant document chunks from large volumes of unstructured data, and Cohere’s embedding model supports semantic understanding.
The chatbot solution uses Llama to craft responses to guest queries in natural language.
THEIR SUCCESS
Streamlined operations and elevated guest experiences
With OCI, Llama and their FRED assistant platform, Brain4Data developed a fully operational chatbot in just two months. Today, the chatbot handles the majority of customer queries for Roatel. The result is reduced workloads for service staff, operational efficiencies for the business and enhanced experiences for customers. With its multilingual AI chatbot, Roatel is now set to scale its micro-hotel network across Europe.
- 80% of customer service requests are handled by Llama-powered AI
- 100% multilingual — the agent converses in the customers’ language
“With no fine-tuning, Llama provides excellent responses to challenging — not always well formulated — questions, in multiple languages.”
Jan-Peter Meyer, CTO, Brain4Data
“OCI Generative AI is a robust, flexible AI development and training platform we used to quickly add Llama-powered AI capabilities to the Brain4Data chatbot. After just a few months, it was able to answer more than 80% of customer questions with a high degree of accuracy, providing customized information and context to customers, just like an employee would.”
Jan-Peter Meyer, CTO, Brain4Data
Previous slide
Next slide
Start building
Explore more
ConsumerBusiness
Shopify | Llama case studiesShopify uses Llama to generate product pages, localize content, and automate support, helping developers scale workflows and save time.Read more
ConsumerTech
Scribd, INC | Llama case studiesDelivering faster, cheaper and more accurate results with a Llama-powered AI content discovery assistant.Read more
Stay up-to-date
Our latest updates delivered to your inbox
Subscribe to our newsletter to keep up with the latest AI updates, releases and more.