Build once, use many: a reusable RAG boilerplate for large organizations

• What is and why do RAG

• Building and Deploying RAG IaaS

• Reusable and customizable components

• Getting and maintaining data for the LLM

• Quick start use case onboarding formula

• Common boilerplate frontend (Streamlit)

• Reusable observability and monitoring

Eric Joachim Liese

Strategic Advisor and Lead Architect for AI & Data
BSH Home Appliances Group

See Bio

Saif Addin Ellafi

Product Lead, MLOps, Data Engineer, Solutions Architect
BSH Home Appliances Group

See Bio

• Exploring LLMs for product portfolio management

• Leveraging LLMs in operational situations to drive

 product and service improvements

• Improving manufacturing through the use of AI

 tools

Iliana Portugues

VP Innovation Centre Berlin
Siemens Energy

See Bio

• Smol LMs are ideal for today’s applications, delivering efficiency without the complexity of massive models; for example, handling routine tasks like summarizing emails or generating quick responses doesn’t require the power of GPT-4.

• Smaller models offer faster processing times and lower costs, making them ideal for edge devices and real-time applications. They also reduce energy consumption and enable on-device operation, enhancing privacy.

• There are ways to enhance these models for better performance, and they’re far more accessible for individual developers and small businesses compared to larger models.

Elie Bakouch

AI Research Engineer
Hugging Face

See Bio

• From idea to action - how to develop an AI-related idea into a workable business solution? Here we want to show how different stakeholders were engaged from the scratch, to ensure smooth business implementation

• From action to proof of concept - how to spend as little as possible on testing what the technology can deliver -  building simple POC with AI Agents

• From mid to big - how to scale AI solutions to ensure best ROI and effective impact across all stakeholders included

• AI Act and GDPR - how to manage regulatory challenges in projects utilizing AI technology

• Small models as foundation for sustainable AI development

Dino Metaxas

Chief Digital & Marketing Officer
L’Oréal

See Bio

Tomasz Chyrchel

AI & Innovation Expert
L’Oréal

See Bio

• Data analysis of patient needs and real language

 used in specific disease areas

• Develop a Gen AI based solution that leverages

 insights and can cross check patient education

 drafts and advise on better language

• Next step: let solution create customized material

 from scratch

Dimitri Metzger

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

See Bio

GenAI for Capital Markets: Are Chat Bots and Smart

Summaries the only solutions?

• The future of Chat Bots for supporting capital

 market research and portfolio construction

• Do Smart Summaries really help the professional

 analyst?

• Is there any potential for more use cases with

 GenAI?

Dieter Konrad

Expert Capital Market Data Science
Union Investment

See Bio

• What is the current art-of-the-possible with regards to Gen AI BI (what it can do and what it’s still missing)

• What are the most often encountered problems when Gen AI BI is introduced in the organization

• How to mitigate possible risks, biases and other challanges that you will encounte

Kazik Surala

Head of Data Governance & Analytics- Poland
Phillip Morris International

See Bio

• Ethics and AI in aviation

• Harnessing the potential of AI while prioritizing

 responsibility

• Ethics by design approaches

Sergei Bobrovskyi

AI Expert
Airbus

See Bio

• Why GenAI is a Big Deal in Legal: How AI is transforming legal work and where it’s making the biggest impact.

• The Legal-Specific Challenges: Why legal processes pose unique hurdles for GenAI.

• What’s Working (and What’s Not): Real examples of GenAI success in legal, and where it’s falling short

David Malkinson

Senior Managing Director Digital Transformation
Morae Global

See Bio

Rico Burnett

Managing Director & Head of Innovation
Morae Global

See Bio

• Understanding Deepfakes: Overview of the technology behind deepfakes, including how they are created and the role of Generative AI in their development.

• The Dual Impact on Society: Examination of how deepfakes are reshaping societal perceptions, from influencing public opinion to the risks of misinformation and identity theft.

• Business Implications: Analysis of both the risks and opportunities for businesses, including the threat and the potential for innovative marketing strategies.

• Ethical and Legal Considerations: Exploration of the ethical dilemmas posed by deepfakes, alongside current and emerging legal frameworks designed to combat malicious uses of deepfakes.

• The Future of Trust: Discussion on the long-term implications of deepfakes for the future of trust in digital content, and how society can navigate the balance between innovation and security

Anna Hovsepyan

Visiting Lecturer
King’s College London

See Bio

• Introduction to Generative AI for Cyber Security

• Using AI for Threat Detection

• Implementation of Security Measures

• Enhancement: Which Tools Support It?

Reserved for Partner

Robert recently joined Newcross to help the business maintain it’s competitive advantage in the age of AI. His first project has been to leverage Generative AI to increase the operational efficiency of the business and effectiveness of its people. In this session Robert will cover:

• Developing a ‘NewcrossGPT’ to enable staff to leverage Generative AI capabilities with internal data and IP

• Creating unique contexts for each team, such as sales, marketing and engineering, to enhance the effectiveness

• The method for deciding what data to train the model on and the governance structures around data access

• Ensuring that data used in the system is both secure and not ending up in OpenAI’s data set

• Taking an agile approach to technology implementation – move quickly and avoid investing too much in any single system given the pace of change

Robert Chilvers

Chief Data & AI Officer
Newcross Healthcare

See Bio

• Putting together a promising-looking prototype of multi-agent AI functionality on a local machine is easy, especially if you use one of the popular frameworks, such as LangChain or LllamaIndex. But how do you deploy and scale it in production? Do you need to use the corresponding framework’s paid offerings, such as LangServe and LangSmith?

• This talk will argue that AI agent orchestration is just a special case of distributed and event-driven workflows, which have been around for a long time. We will show how, for example, LlamaIndex’s Workflows semantics can be mostly replicated with Faust and Kafka; and briefly go over the many open source alternatives to LangSmith.

• Finally, we will describe the way Wise is going about productionizing and scaling its LLM workflows using Ray, a generic Python scalability framework.

Dr. Egor Kraev

Head of AI
Wise

See Bio

• Explore the complete journey of Generative AI development, from ideation to production.

• Discover impactful use cases and learn best practices for implementing Generative AI

 solutions.

• Understand strategies for leveraging shared services to enhance AI capabilities and efficiency

• Learn about Red Teaming methodologies to proactively identify and mitigate vulnerabilities in

 AI systems.

• Delve into the critical aspects of AI security to protect your AI systems from threats.

Ömer Adıgüzel

Technical Lead Generative AI
Lufthansa Group

See Bio

This thought-provoking talk explores the unexpected ways AI is transforming our world, challenging our preconceptions

about the future of work, personal relationships, and human capabilities. Through interactive discussions and real-world

examples, we’ll delve into the paradoxical nature of AI’s impact on society and business.

Stiv Sterjo

Smart Analytics & AI Cluster Lead
Google

See Bio

• Evaluating Copilot M365: Investigating how Microsoft’s generative AI tool, Copilot M365, influences efficiency, quality, and employee experience within Repsol.

• Experimental Study: A four-month research with 550 participants, using pre and post-test design, practical experiments, surveys, focus groups, and interviews to measure the impact.

• Key Findings: Results show a weekly time saving of 121 minutes, a 16% increase in quality of deliverables, and a significant positive impact on employee satisfaction and productivity

Dr. Guille Lorbada

Head of New Ways of Working
Repsol

See Bio