Geodis operates a global logistics network, supported by a corporate IT service desk that assists thousands of employees worldwide. Every day, IT teams must quickly understand issues, route tickets to the right experts, and find accurate resolutions — all while meeting strict service-level expectations.

In practice, this wasn’t easy. While many tickets were relatively simple, only about half were resolved at Level 1. Misclassification and unclear working instructions meant tickets were frequently escalated to Level 2 or Level 3, even when they could have been handled earlier. Valuable resolution knowledge existed across past tickets and documentation, but it was difficult to find, reuse, and apply consistently.

Geodis wanted to improve efficiency and service quality without overhauling existing systems or replacing human expertise. The goal was to reduce manual effort, speed up resolution, and give IT teams better support in their day-to-day work, while building a foundation that could scale across IT operations over time. To achieve this, they turned to the Platform for AI Success from Dataiku to design, deploy, and scale AI agents within their existing workflows.

Results with Dataiku

By embedding an AI agent directly into IT support workflows — bringing orchestration across data, machine learning models, and applications like ServiceNow — Geodis transformed how tickets are assigned and resolved. This enabled them to move from manual triage to faster, AI-assisted support, helping teams work more efficiently while staying in control of how issues are handled.

Faster assignment, less manual effort: Ticket classification and assignment, previously a time-consuming manual step, is now automated. Early estimates show a 60% reduction in assignment time, immediately removing friction at the start of the support process.

Quicker resolutions, better focus: With tickets routed correctly and relevant guidance surfaced upfront, IT teams save around 30 minutes per ticket on average. Support agents spend less time searching and reassigning, and more time resolving issues, especially at Level 1.

Faster IT ticket resolution with an AI IT Support Agent

Geodis built an AI IT Support Agent in Dataiku to assist support teams directly within ServiceNow, without forcing them to adopt new tools or workflows.

For each incoming ticket, the agent:

• Classifies the issue and predicts the correct support level and resolution group

• Identifies similar historical tickets

• Suggests resolution steps using internal technical documentation and past cases

All recommendations appear inside the ServiceNow interface, where members of the support team can review and apply them with a single click. The AI agent augments human expertise rather than replacing it, empowering support personnel to remain fully in control and decide how to act on the recommendations.

The impact is immediate and practical. Tickets are routed more accurately, unnecessary escalations are reduced, and Level 1 teams are able to resolve a higher share of issues. As a result, IT support work becomes less repetitive and more engaging, while internal users experience faster, smoother resolution.

Beyond

productivity gains, the agent also improves day-to-day responsiveness. By making knowledge easier to access and reuse, teams can react more quickly to recurring issues, reduce manual bias in triage, and spot patterns that were previously buried across tickets and documents.

The Dataiku difference

Geodis chose Dataiku because it provides a single platform to design, deploy, and govern AI agents— while keeping people at the center of the process.

Key differentiators include:

• Human-AI collaboration by design, where agents support decision-making and execution without removing human oversight

• No-code to full-code workflows, allowing data scientists, IT experts, and analysts to collaborate in one environment

• Seamless integration with existing systems, embedding AI directly into ServiceNow rather than adding another tool

• Enterprise-grade governance

  , to monitor performance, cost, and risk across AI systems — ensuring AI outputs remain controlled, auditable, and scalable

Vision & what's next

Geodis is using this AI IT Support Agent as a blueprint for scaling AI across IT operations. The approach is intentional: Start with a focused, high-impact use case, validate results, and expand once trust and performance are proven.

Next steps include rolling the solution into production for the full corporate service desk, increasing the share of tickets resolved at Level 1, and continuously improving the underlying knowledge base and working instructions. Longer term, Geodis sees this agent as a stepping stone toward more advanced automation, including AI-driven “Level 0” support, while maintaining a strong human-in-the-loop model.

By consolidating analytics, machine learning, and AI agents in Dataiku, Geodis is building a governed, scalable foundation for AI success — enabling teams to operationalize AI agents across the business and work smarter, faster, and with greater impact.