How Agentic AI is Revolutionizing Collaboration at Tchibo

Tchibo GmbH is one of Germany’s most recognized consumer goods and retail companies. Beyond its core coffee business, the brand is renowned for a weekly rotating line of non-food products. To stay agile in a dynamic retail environment, the long-standing Hamburg-based enterprise increasingly relies on innovative technologies, modern data architectures, and targeted artificial intelligence applications.

The project at a glance

Design and continuous evolution of a central agent platform for internal employees
Development of a specialized data analysis agent to streamline queries across complex data sources
Integration of key business tools and document sources via standardized MCP servers

Targeted AI enablement initiatives and interactive formats for internal knowledge sharing
Highly agile approach within a dynamic, fast-evolving market and tech ecosystem

Background

Continuous advances in Generative AI open up massive potential for greater efficiency and smarter business processes. Tchibo saw a growing desire across departments to integrate innovative AI tools into their daily workflows. Recurring evaluations, writing queries for complex data structures, and preparing monthly reports were consuming significant time and specialized expertise across many teams.

At the same time, companies face the challenge of an unprecedented pace of AI development. Market standards, models, and interfaces are constantly shifting. Rigid development frameworks quickly hit their limits, as newly emerging technologies can swiftly render existing structures obsolete. Tchibo needed a way to provide employees with a secure, flexible, and scalable AI platform that could keep pace with market trends and simplify everyday work.

Solution

In close collaboration with Tchibo, an expert team from codecentric AG developed an agile agent platform for internal use. Through a centralized chat interface, employees gain access to various modern Large Language Models (LLMs) and specialized assistants—ranging from data analysis and personal knowledge management to image processing and corporate-design presentation creation.

A core focus of the partnership is the development of a specialized Data Analyst Agent. This tool empowers employees to query and evaluate complex data and related documentation using natural language.

Flexible Architecture & Tool Integration: Data sources and existing systems—such as OneDrive, SharePoint, Confluence, Jira, BigQuery, DataHub, Censhare, and external sources like the SERP API—are connected via MCP (Model Context Protocol) servers. This unlocks broad application possibilities for internal knowledge management and market monitoring.
Close Departmental Collaboration: Because the Data Analyst Agent relies on meticulously maintained context and data structures, the tech team works closely with the respective business units. Through targeted context engineering, they established the foundation for the agent to ask the right questions and deliver precise results. The departments bring their domain expertise to the table and continuously refine the necessary documentation.
Agile and Adaptive Approach: To respond instantly to new market and tooling developments, the project is built on a highly flexible development model. In-house developments are continuously evaluated and, when appropriate, replaced by high-performing external solutions or expanded interfaces.
Culture & Enablement: Beyond technical implementation, the team supports rollout through interactive formats and training sessions, such as internal hackathons. This builds a solid understanding of both the potential and limitations of the technology across the entire company.

Result

With the launch of the agent platform, Tchibo has successfully fostered widespread adoption and growing usage of generative AI across the company. Employees from various departments leverage these AI agents to streamline daily routines and complete task workloads significantly faster.

The Data Analyst Agent enables teams to extract deeper insights from existing data pools without requiring specialized database query expertise. Today, use cases range from e-commerce KPI analysis, product line performance, and conversion tracking to campaign sales evaluations, inventory and replenishment tracking, and AI-driven shelf availability in physical stores. Each department embeds its own domain logic, configuring the agent for the specific data and KPIs critical to its daily operations. This makes sharing data and analytical insights across the company far simpler and more transparent. Thanks to this agile setup, the platform remains permanently adaptable—building a strong foundation for future retail AI use cases.

Dominik Nowatschin

Lead AI Platform & Engineering, Tchibo GmbH

The dynamic evolution of the AI space demands maximum flexibility. Thanks to our collaborative and agile partnership with codecentric, we’ve successfully built a platform that directly meets our needs and evolves alongside the latest technologies.

Dominik Nowatschin

Lead AI Platform & Engineering, Tchibo GmbH

Any questions about the project?

Would you like to unlock the full potential of complex data across your departments and empower non-technical teams? Let’s connect! Together, we’ll explore how custom AI assistants can streamline your daily operations.

Janine Felten