Main Conference Day 1, Wednesday, 4 December 2025

• Overview of the current state of AI integration in shared

service centers, highlighting practical use cases and pilot

projects.

• Exploration of key challenges faced during AI

implementation, including ROI measurement, change

management, and technology adoption hurdles.

• Insights into future trends and strategic priorities for scaling

AI-driven automation across global business services.

Marc Sauty

Transformation Director
Sodexo

See Bio

• Use of orchestration frameworks to chain LLM calls with tools,

APIs, and memory for complex multi-step workflows.

• Case studies from document processing, procurement, and

service resolution with fully or semi-autonomous agents.

• Alignment with platform strategy, change management, and

IT compliance for controlled scaling of agentic systems.

Slot reserved for one of our business partners

• Exploration of practical use cases of generative and agentic

AI transforming back-office and global business services

operations.

• Discussion of key application areas including finance, HR,

procurement, and compliance within GBS environments.

• Vision for the future role of autonomous AI workflows in

driving efficiency, scalability, and strategic innovation across

back-office functions.

Marc Sauty

Transformation Director
Sodexo

See Bio

• Use of agentic systems to search, retrieve, and synthesize

scientific literature in response to medical information queries

or research hypotheses.

• Deployment of a modular platform that integrates licensed

databases and external tools to validate or disprove scientific

assumptions.

• Application of autonomous agents to orchestrate search,

ranking, summarization, and evidence tracking in regulated research workflows.

Dimitri Metzger

Digital, Data & IT Business Partner for Development & Medical Units & AI
Merck

See Bio

• Showcase of real-world agentic AI applications across

enterprise functions such as customer service, supply chain,

and finance.

• Overview of deployment approaches, including integration

with existing IT systems and overcoming operational challenges.

• Best practices for scaling agentic AI solutions, focusing

on governance, performance monitoring, and continuous

improvement.

Slot reserved for one of our business partners

• Outlines the transition from manual capital market research

to AI-supported decision-making using Deep Research tools.

• Highlights regulatory, cultural, and data licensing barriers to

deploying agentic AI in financial institutions.

• Explores the vision of AI-integrated team collaboration and

real-time decision support through conversational interfaces.

Dieter Konrad

Expert Capital Market Data Science
Union Investment

See Bio

• ASN Bank is using Agentic AI to assist with unifying four

brands

• An agent to trace data lineage across legacy systems to

provide quick insights and aid migrations

• A policy agent analyzed and compared policies to create

a common glossary and identify duplication or overlap in

definitions

• The architecture and design patterns behind these AI agents

will be discussed

Steven Laan

Lead Data Scientist
de Volksbank

See Bio

• Navigating the shift from analytics-led to data-first thinking

in enabling sustainable AI transformation.

• Building a data quality framework that supports compliance,

scalability, and model accuracy across a global enterprise.

• Lessons from Ericsson’s AI adoption journey: organizational

change, governance, and aligning architecture with business

needs.

Gabor Harsanyi

Head of M&CR Data Office
Ericsson

See Bio

• Implementation of modern data Lakehouse architectures

and vector databases to support retrieval-augmented

generation, semantic search, and real-time AI applications.

• Adoption of hybrid storage models combining structured,

unstructured, and time-series data sources for GenAI use

cases across manufacturing, pharma, and logistics.

• Approaches for data orchestration, lineage tracking, and AI-

ready metadata management to ensure enterprise trust in LLM

outputs and data traceability.

• DT position on the EU AI Act: GPAI models and harmonized

standards

• How to implement it in DT

• Extension to Green AI

Joachim Stegmann

Chapter Team Lead Artificial Intelligence & Data Science
Deutsche Telekom AG, Innovation Hub

See Bio

• How enterprises translate fairness, transparency, and

accountability into enforceable technical and organizational

measures across the AI lifecycle.

• Explore methods for risk classification, traceability, and

human oversight aligned with internal governance and

external regulation (e.g. EU AI Act).

• Review best practices in model documentation, auditing,

and cross-functional collaboration between legal, IT, and data

science teams.

• Guidance and quick wins in the journey towards Sustainable

Software&AI maturity in teams and organizations

• Examine requirements for aligning AI product development

with upcoming EU sustainability initiatives and regulatory

frameworks.

• Explore practical approaches to start measuring and

reducing the energy footprint of AI systems

• Use of GenAI to tailor content, product recommendations,

and promotions dynamically across digital and physical retail

channels.

• Integration of customer data platforms, behavioral analytics,

and retrieval-augmented generation (RAG) to generate

individualized messages and offers in real time.

• Implementation within enterprise marketing workflows,

balancing privacy compliance (e.g. GDPR) and performance

measurement across customer journeys

• Deployment of autonomous AI agents to coordinate adaptive

targeting, HCP engagement, and personalized content delivery

across channels.

• Integration with CRM systems, medical content libraries, and

compliance workflows to ensure traceability, brand safety, and

regulatory alignment.

• Use of reinforcement learning and retrieval-augmented

architectures to continuously improve performance based on

user feedback and behavioral signals.

• Enable marketing teams to generate audience-specific

content at scale by combining creative prompts with real-time

behavioral data.

• Leverage GenAI to uncover hidden patterns in customer

interaction and inform campaign strategies with adaptive

insight.

• Integrate human creativity and AI-driven ideation in co-pilot

models that preserve brand voice while increasing output.

• Adoption of compact, parameter-efficient models enables

cost-effective and privacy-conscious GenAI deployments

across enterprise domains.

• Techniques like LoRA, distillation, RLHF, and RAG architectures

support lightweight fine-tuning on proprietary data.

• Use cases such as internal chatbots, summarization tools,

and personalization engines benefit from on-premise, domain-

specific AI integration.

• Connecting LLM-powered agents to internal systems through

APIs to automate tasks across domains such as customer

service, compliance, and operations.

• Introduction of middleware strategies, API wrappers, and

lightweight orchestration layers to ensure robust and secure

AI-to-system interactions.

• Governance and team enablement challenges in exposing

enterprise functionality to GenAI tools while maintaining control

and compliance.

• Use of simulation environments and synthetic data

generators to train and validate models in constrained or high-

risk domains.

• Applications in digital twins, predictive maintenance, and rare

case generation in pharma, manufacturing, and mobility.

• Evaluation frameworks to compare synthetic to real-world

datasets for performance, bias, and regulatory acceptance.