ARCHITECTING CUSTOM AI MODELS AND SCALABLE INFRASTRUCTURE FOR ENTERPRISE ADOPTION AND VALUE GENERATION
January 14th-15th, 2025 |
Hilton London Syon Park
Main Day 2 - Wednesday 15th January 2025
8:35 am - 8:40 am
Chair Remarks
Maria Stefanova -
Head of PMO – Artificial Intelligence,
International Airlines Group (Iag)
8:40 am - 9:20 am
PANEL: Integrating AI into Organisational Workflows and Fostering Cross-Business Collaboration
Trevor O'Brien -
Senior Director, Technology Innovation,
Moody's Analytics
Ahmed Elahi - Senior Director of Data, SkyShowtime
Jie Zheng - Technology Team Lead - Machine Learning Lab, TUI
Shruti Kohli - Head of Data Science, Innovation and AI, Innovation lab, Department for Work and Pensions (DWP)
- Establish an operating model to integrate AI products into business applications without losing their value.
- How to transition a project from innovation (0-1/risk taking teams) to operation (1 - production / risk neutral teams) to systematic / long-term project (1 – N / risk averse teams). The core value of the delivery team for each phase of the project shift and changes. How recognising and respecting the value and differences between stages and teams can boost efficiency of collaboration and improve overall impact of AI projects.
- Securing Internal Buy-In and Resources: Cultivate support across the organisation by clearly communicating the value of the AI product strategy, ensuring stakeholders understand its benefits and are committed to investing the necessary resources for successful implementation.
- Team collaboration and cross-function innovation: Interactions between product, data science, engineering, and user experience teams to streamline the development and launch of AI applications. Enabling these teams to innovate and scale AI initiatives, integrating AI capabilities into business processes.
Shruti Kohli
Head of Data Science, Innovation and AI, Innovation labDepartment for Work and Pensions (DWP)
9:20 am - 9:50 am
Scaling AI: From Concept to Enterprise Impact
Jon Bratseth -
CEO and Founder,
Vespa.ai
- Getting AI Out of the Lab: barriers to scaling AI from prototypes to enterprise-wide deployment—high costs, complexity, diverse workloads, data management, and performance—and strategies to overcome them.
- Simplification with AI Platforms: how integrated platforms streamline AI development, deployment, and management by reducing complexity, boosting efficiency, and enabling teams to deliver impactful solutions.
- Emerging Technologies: Driving the Future of AI: key advancements like Vision-Language Models and technologies such as PDF document embedding and retrieval, unlocking new opportunities for enterprise AI innovation and cost reduction.
9:50 am - 10:30 am
PANEL: Delivering Enterprise Specific AI Workloads with Cost Effective and Efficient Infrastructure
Richard Kiernan -
Global Head of AI & Machine Learning Platforms,
Natwest RBS Group
Cameron Farrar - VP, Head of Software Asset Management, Marsh & McLennan Companies
Rami Mäkelä - Head of Data Architecture, Intellectual Property Office UK
David Henstock - Chief Data Scientist, BAE Systems Digital Intelligence
- Ensuring AI-Ready Data: Learn how to manage and prepare high-quality data to power AI models effectively, ensuring optimal results in dynamic and data-rich environments.
- Successful AI deployments, focusing on the infrastructure strategies that enabled businesses to boost operational efficiency, enhance productivity, and drive revenue growth while reducing costs.
- Optimising AI Infrastructure for Business Demands: Explore strategies to manage increasing computational requirements and overcome challenges like latency, power consumption, and compute access, with a focus on cost-efficient, scalable solutions that align with business objectives.
- Enhancing AI Deployment with MLOps: Streamlining AI model lifecycle management, optimizing infrastructure, reducing costs, and improving scalability and reliability in enterprise settings.
10:30 am - 11:00 am
Networking & Refreshments
11:00 am - 11:05 am
Chair Remarks
11:05 am - 11:30 am
Optimising Small Language Models (<= 7B parameters) on commodity hardware to address different requirements (costs, privacy, safety, regulations, etc.) in the BioTech Manufacturing
Guglielmo Iozzia -
Director – Data Science, ML/AI, Computer Vision,
MSD
11:30 am - 12:00 pm
SPONSORED SESSION: Enhancing AI Efficiency: Navigating Hardware and Software Solutions for Optimized AI Workloads
- Addressing Scalability and Performance Challenges: Explore strategies to balance cost, power consumption, and performance when deploying AI on constrained hardware.
12:00 pm - 12:25 pm
Creating Value and Solving Real-World Problems by Applying AI and Edge Computing
12:25 pm - 1:00 pm
PANEL: Transforming AI Workloads with Scalable Hardware and Software Integration
- Overcoming AI Architecture Challenges: Address scaling issues in distributed systems and the architectural shifts needed for AI-driven demands.
- Leveraging Hardware Acceleration: Explore how FPGAs, ASICs, and specialized processors enhance AI performance and energy efficiency.
- Maximising Hardware Utilization: Learn techniques for optimizing software to fully exploit advanced GPUs and processors for efficient AI training and inference.
- Aligning Software and Hardware: Discover how synchronizing development cycles can overcome bottlenecks, improving scalability and efficiency in AI deployments.
1:05 pm - 3:00 pm
Lunch
2:00 pm - 2:05 pm
Chair Remarks
David Gerouville-Farrell -
Chief AI Officer,
Stealth Mode Startup
2:05 pm - 2:30 pm
Strategic and Responsible Development of AI Products: Balancing Innovation with Architecture, Design & Ethical Considerations
Mayank Srivastava -
Global Director Data and Analytics,
Frieslandcampina
2:30 pm - 3:00 pm
SPONSORED SESSION: Leveraging Multimodal AI Models with Advanced Hardware for Enhanced Business Deployments
- Integrating Data Modalities: Explore the benefits of combining text, images, and audio within AI models to improve accuracy and context-awareness in business applications.
- Discuss emerging advancements in multimodal AI models and the hardware innovations that will drive further performance improvements and business innovation.
- Optimised Deployment Strategies: Learn best practices for deploying multimodal models, addressing challenges in data integration and resource management.
3:00 pm - 3:25 pm
From Research to Reality: Innovating and Creating User Centric AI products
Martina Johannesson -
Senior Technical Product Manager,
Unity Technologies
Quentin Theillaud - Senior Manager, Machine Learning, Unity Technologies
3:25 pm - 3:50 pm
Scaling Impact: Save the Children's AI Product Development Strategy
Nicholas Drabowski -
Head of Generative AI Workstream,
Save the Children
- Cost-Effective Resource Management & Donor Alignment: Leverage AI infrastructure to maximize resource efficiency, aligning with third-party donor expectations while minimizing costs.
- Eliminating language barriers in RAG knowledge search for Global Operations: Develop AI systems that handle diverse languages effectively, ensuring that solutions are accessible and functional across all regions.
- Achieving Scalable Product Solutions: Design AI-driven products that can be scaled globally, focusing on maintaining cost-efficiency, reliability, and adaptability to diverse environments.
3:45 pm - 4:10 pm
Networking & Refreshments
4:10 pm - 4:15 pm
Chair Remarks
David Gerouville-Farrell -
Chief AI Officer,
Stealth Mode Startup
4:15 pm - 4:40 pm
How to Integrate AI Products and Lead Transformation of a Business Operating Model
Simon Barna -
Programme Director, Data & Ai Area Product Manager,
BT
4:40 pm - 5:20 pm
PANEL: Enabling AI-Ready Data Platforms and Build Valuable AI Products
Omololu Makinde -
Senior AI Solutions Engineer,
Ofcom
Nikita Iserson - Director, Machine Learning Engineering, S&P Global
Aleena Annie - AI, Modelling and Insights Lead, Lloyds Banking Group
- The critical role of high-quality, actionable data in powering AI and generative AI (GenAI) initiatives.
- Strategies for safeguarding data as a key enterprise asset, ensuring its integrity, privacy, and security while maximizing value from AI-driven insights.
- Ensuring data is findable, accessible, trustworthy, interoperable, and reusable to enhance AI outcomes.
- Implementing and enforcing robust data governance policies through advanced data platforms.
- Exploring the future of business models shaped by the integration of AI into products and services.
