• Tools and techniques for model selection, training, and customisation, including foundation models, fine-tuned models, and centralised repositories to streamline AI development.
• Deploying and Orchestrating AI/ML Models: How to effectively deploy, distribute, and orchestrate AI/ML models to ensure your solution is accurate, reliable, and scalable. Considering specific requirements for training and inference when designing AI/ML solutions. How to establish and meet performance benchmarks for different components of your system, ensuring optimal operation.
• The importance of a robust infrastructure layer, including cloud platforms and specialized hardware (GPUs, TPUs), to support the intensive training and inference workloads of generative AI models.
• How platform engineering for MLOps enhances the orchestration of AI models, focusing on adapting, deploying, and monitoring models effectively within end-user applications.

Andre Dourson

Next Generation Technologies Director
MARS Petcare

See Bio

Yeunjin Kim

Head of Data Science
Mott Macdonald

See Bio

Hamza Tahir

Co-Founder
ZenML

See Bio

Enterprises adopting generative AI face a trade-off: maintaining high precision (16-bit) for trustworthy results at prohibitive costs, or compromising quality with low precision (e.g. 4-bit) to manage expenses. Quantization, while cost-effective, can introduce quality losses in average and outlier performance, which are difficult to evaluate in generative AI (e.g. media gen, or trading systems). This dilemma hampers adoption in critical enterprise applications, where trust in AI results is paramount - as enterprises are moving towards AI as co-pilots and pilots.

Recogni introduces Pareto, a groundbreaking AI math approach leveraging logarithmic scales to eliminate costly multiplications—transforming them into simple additions at no loss in precision. Pareto accelerates models in true 16-bit, at the cost of running it in 4-bit on available systems, enabling enterprises to scale generative AI affordably while retaining high-quality outputs.

This presentation explores Pareto, and demonstrates how it unlocks enterprise adoption of GenAI by bridging the gap between trust and affordability.

Gilles Backhus

Founder/VP of AI and Product
Recogni

See Bio

• AI-ready data: Ensuring quality, structure, and governance to build reliable AI models.
• Best practices: Industry best practices for designing and refining custom models.
• Competitive advantage: How tailored AI models drive impact; providing a unique edge and measurable ROI in enterprise settings.
• Use case identification: Define and prioritize high-impact, business-relevant AI use cases that align with strategic goals.

Dara Sosulski

Head of Artificial Intelligence and Model Management
HSBC

See Bio

Ali Faisal

Lead AI product development
Coca-Cola Europacific Partners

See Bio

Jürgen Weichenberger

Former VP of AI Strategy & Innovation
Schneider Electric

See Bio

Jeremy Smith

Chief Technologist, Incubation Projects in Advanced Compute & Solution
HP

See Bio

The objective of the talk is to outline that the ongoing business value is derived from operationalizing AI by consistently synchronising along four concurrent dimensions: Infrastructure, Data, Code and Model.

Each of these dimensions have traditionally different stakeholders with different imperatives, which create friction whilst bringing models from labs to live production.

ML/AI mandates to look at these 4 dimensions concurrently, consistently and coherently along the exploration, development, testing, validation and production lifecycle.

Amit Nandi

VP Solutions & Data Architect
Barclays Corporate & Investment Bank

See Bio

• Session will focus on how data is key to the success of any AI implementation.
• “Good data” – what does this mean and how do you get “good data”
• “Lots of it” - will look at how more data provides better learnings. How do you gather and create data and how do you manage such vast reserves of data resources
• “Fast” - Extreme high speed access to data are table stakes for any data storage platform to operate successfully in AI infrastructure environments. Why is this and how do you achieve TBps data transfers?

Neil Stobart

CTO
Cloudian

See Bio

• Scalability and delivery of AI custom workloads through specialised infrastructure, enabling rapid deployment and scaling of AI models across distributed environments.
• Insights into the capabilities necessary to deliver a custom compute platform that meets the demanding requirements of AI, ensuring optimized performance and efficiency.
• How hyperscalers are enhancing AI infrastructure, providing scalable, high-performance resources that can support vast AI workloads while maintaining flexibility and cost-effectiveness.

Shelly Henry

Director of Hardware Engineering
Microsoft

See Bio

Kamran Naqvi

Chief Network Architect – EMEA
Broadcom

See Bio

Patrick Lastennet

VP Strategic Partnerships
OpCore - Group Iliad

See Bio

Renato Paço

Global Director of Cloud and Infrastructure
Sky Group

See Bio

• Machine learning lifecycle and challenges in scaling ML products 

• Core components of machine learning platform and best practices enforcement

Hisham Mohamed

Director of Engineering, Machine Learning Platform
Expedia Group

See Bio

• Critical Components for Success: Process, People, and Technology: A deep dive into the three essential pillars of AI product development: optimizing processes, aligning skilled teams, and leveraging the right technologies for successful deployment.
• Enabling Scalable Enterprise AI Deployment: Strategies to implement AI solutions at scale, focusing on performance, reliability, and efficiency for large-scale enterprise environments.
• Enterprise-wide AI Integration Best Practices: Insights on integrating custom AI models across diverse platforms, focusing on scalability, consistent performance, and streamlined workflows. These practices help accelerate AI-driven innovation throughout the organization.
• Optimizing Scalable AI Infrastructure: Explore how to design infrastructure that supports seamless deployment of AI models by leveraging cloud-based platforms, distributed computing, and hybrid solutions. Learn to balance cost, performance, and operational efficiency, with a brief focus on how MLOps can play a role in ensuring scalability across the enterprise.

Donatien Chedom Fotso

Head of Applied AI – CB Data Science Center of Excellence
Deutsche Bank

See Bio

Tharun Loknath

Head of AI Architecture
Just Group

See Bio

Bugra Ozer

Data Science and AI Governance Lead
Dentons

See Bio

Eliza Upadhyaya

Product Manager
deepset

See Bio