eXplainable AI Day 2024
How to Explain and Report Decisions Made by Artificial Intelligence?
Welcome to eXplainable AI Day 2024, the first and unique event in Poland dedicated entirely to explaining the decision of AI models. We invite you to join us during our conference to gain cutting-edge insights into AI, learn techniques for extracting information out of AI models, and explore practical use cases.
5th March 2024
Agenda
10:00-10:05
Event opening
10:05-10:40
KEYNOTE: The Purpose and Value of Explainable AI
Explainable AI is more than just a technical necessity. This keynote will discuss the significance of explainable AI (XAI) beyond its conventional role in model improvement and debugging. Explainable AI plays a profound social role as it impacts trust and accountability. Our understanding of XAI, as a consequence, might determine the success and long-term legitimacy of our organization, public and private.
Joris Krijger, PhD
AI and Ethics Officer at De Volksbank
10:40-11:05
Regulating explainable AI
AI without any doubt will be the technology of tomorrow, and some would say that it is a technology of today. Essential features of explainable AI – its need for transparency, interpretability, and accountability are safeguarded in draft reguations. We explored how regulations should guide AI to be comprehensible to its users, ensuring decisions made by AI systems are clear and justifiable. This approach is crucial for building trust and facilitating ethical AI applications in our increasingly digital world.
Marek Porzeżyński, PhD, CIPP/E, MBA
Partner, Head of IP and NewTech at the Linke Kulicki Law Firm
11:05-11:30
AI regulatory landscape – what’s now and what’s next. How to turn compliance into your competitive advantage
In an era where artificial intelligence (AI) is transforming industries and everyday life, understanding the AI regulatory framework and leveraging compliance for competitive advantage is crucial. The presentation will offer a comprehensive overview of the current AI regulatory environment, highlighting key legislation and ethical considerations critical for developers and businesses. We’ll explore how jurisdictions worldwide are shaping AI development and deployment, emphasizing the importance of transparency, accountability, and fairness in AI systems. More importantly, we will delve into how regulatory compliance is not just a necessity but a strategic asset that can differentiate businesses in a competitive marketplace. I aim to demonstrate how companies can transform compliance challenges into opportunities for innovation, trust-building, and market leadership.
Łukasz Węgrzyn
Head of IT and Data Team at Osborne Clarke
11:30-11:50
Coffee Break
11:50-12:15
Role of AI in physical infrastructure
In the energy industry, in the area of IT or OT systems, we deal with large data sets. We use this data directly or after aggregation in the area of business decisions, but also in operational work. It is very important to understand the algorithms, bring at the end to the result, which is an element of further decision-making. In this context, a huge role will be played artificial intelligence, which can give a different speed to the organization. But the key is the explainability of artificial intelligence models.
Mariusz Jurczyk, PhD
Executive Director of Business Continuity Management System, IT and OT at TAURON Polska Energia
12:15-12:40
Machine learning in e-commerce: how artificial intelligence can drive your business
The e-commerce sector in Poland is rapidly developing, establishing our country as one of the global leaders in this industry. In 2023, as much as 80% of internet users made online purchases, which opens up extraordinary opportunities for us, but at the same time poses significant challenges associated with handling such a large scale of operations. In this context, machine learning plays a key role.
Witold Zaklukiewicz
Data Science Team Leader at Ceneo.pl
12:40-13:05
Unleash the Power of Azure AI for Your Business Success
This presentation delves into the AI ecosystem available on Microsoft Azure, highlighting key services such as computer vision, AI speech, and the increasingly popular Azure OpenAI with GPT models. We will explore potential use cases for these technologies within organizations and discuss the essential questions that should be considered when initiating a new AI project. The talk will also cover various approaches to AI solutions, including the use of Retrieval-Augmented Generation (RAG) architecture and the deployment of AI agents. Additionally, we will address the crucial topic of explainability: strategies for managing and monitoring the behavior of AI-based solutions.
Agnieszka Niezgoda, PhD
Senior Technical Specialist, Data & AI at Microsoft
13:05-13:30
Managing rapid progress – how to automate processes within a company in a hyper-agile way?
Whether you want to utilise AI, Low-Code, or RPA in your company, you will face a similar problem: the waterfall approach and even agile don’t work as effectively. What’s needed is an approach that keeps pace with technology, and hyper-agile management offers greater efficiency, better monitoring, predictability, digital security, and regulatory compliance. By leveraging business analysis, IT, and new roles like Citizen Developer, processes can be optimised in as little as 5 days. The session reveals experiences from Budimex, NFM Group, and Tchibo, demonstrating how to achieve success by utilising modern management and technologies.
Mike Guzowski
CEO, Co-Founder of Developico
13:30-13:50
Coffee Break
13:50-14:15
Explaining Machine Learning models for CRM
Machine Learning models used for CRM data allow us to predict patterns and profile customers. Unfortunately, these models are also complex and inherently opaque. This can be overcome by techniques of eXplainable AI. During the talk I’ll present how Shapley values can be used to extract information from Machine Learning models and show typical business use cases for such analysis.
Łukasz Borowiecki, PhD
CEO, Co-Founder of 10 Senses
14:15-14:40
Why does this look like that? Demystifying AI in Computer Vision
Modern AI systems for computer vision are used globally every day, from autonomous driving systems to image generation models. Although tremendously useful in many applications, these systems are not without their pitfalls. In this talk, we will showcase examples of some common errors made by modern image processing models and discuss how integrating interpretability into these methods can lead us toward creating more robust and safer solutions.
Mikołaj Sacha
Head of Machine Learning at Molecule.one
14:45-15:05
Explicit reasoning and explainability in Large Language Models – practical applications for generating documentation
As businesses increasingly rely on Large Language Models (LLMs) and Generative AI (GenAI) systems, the ability to explain their decision-making processes becomes crucial, ensuring the content they produce is not only useful but also clear and transparent to users. In this talk, I will explore how explicit reasoning mechanisms powered by LLMs, along with other explainability frameworks, can help achieve this goal. We will examine practical examples of generating documentation and discuss best practices aimed at enhancing trustworthiness and effectiveness. Additionally, this talk will also cover the challenges of balancing the complexity of these models with the need for the whole process to be understandable and as transparent as possible.
CTO, Co-Founder of 10 Senses
15:05-15:10
Event Closing
Speakers
Joris Krijger, PhD
AI and Ethics Officer at De Volksbank
Joris Krijger is a Tech Ethicist working as Ethics & AI Officer at the Dutch bank de Volksbank while also holding a PhD position on the topic of Ethics and Artificial Intelligence (AI) at the Erasmus University Rotterdam. He presently works on bridging the gap between ethical principles and AI by studying the operationalization of ethics in data science contexts. Additionally, he holds positions as a.o. Founding Editorial Board Member of Springer Nature’s AI and Ethics Journal. For his contributions to the field of responsible AI he received a.o. a Graduate School Award for PhD Excellence.
Marek Porzeżyński, PhD, CIPP/E, MBA
Partner, Head of IP and NewTech at the Linke Kulicki Law Firm
Lawyer and researcher specializing in legal aspects of new technologies, with particular emphasis on intellectual property law, privacy, cybersecurity and AI. Assistant Professor at Warsaw University of Technology, Partner at the Linke Kulicki Law Firm. Graduated MBA – Innovation and Data Analysis program (IPI PAN and Utah Valley University – Woodbury School of Business) and studied applied computer science (WUT). Internal auditor of information security systems and member of The WUT team for the exploration and exploitation of AI capabilities, or the working group for cybersecurity at the Ministry of Digitization. He gained his professional experience in the largest law firms in the world and in Poland. Author of almost 50 scientific works and 2 monographs in his field of specialization.
Łukasz Węgrzyn
Head of IT and Data Team at Osborne Clarke
Expert in legal regulations concerning technology. He specializes in advising highly regulated sectors: finance, insurance, health. Recommended in the international ranking of FinTech Chambers as one of the best Polish lawyers in the field of financial technology. In The Legal 500 and Chambers Europe rankings, he is recommended in the areas of Technology-Media-Telecommunications. Member of the Institute of New Technologies and Personal Data Law at Lazarski University. Member of the New Technologies Law Association. He holds the following technological certifications: ITIL Foundation, PSMI (Professional Scrum Master), AgilePM Foundation, Professional Agile Leadership I.
Mariusz Jurczyk, PhD
Executive Director of Business Continuity Management System, IT and OT at TAURON Polska Energia
Since 2000, he has been associated with the TAURON Group. He was responsible for the AMI rollout in the area of Wrocław city – the first such large-scale urban implementation in Poland. He held various managing positions in areas of advanced metering and distribution. Currently, he is in charge of IT (CIO), OT, Business Continuity Management System, and Digital Transformation at TAURON Polska Energia. He is a member of the Working Group at the Ministry of Climate and the Environment related to smart metering. He is a co-author within PTPiREE of the feasibility study for the implementation of smart metering in Poland. He is also a member of CIONET and the ITAM Forum Polish Chapter. He was in charge of the AMI Working Group at the Polish Power Transmission and Distribution Association at PTPiREE. Additionally, he was a member of the Cybersecurity IT/OT Working Group at PTPiREE and the Working Group for Smart Grid networks at the Central Office of Measures, the Polish Committee for Standardization. He is the author of many publications at conferences related to smart metering, cybersecurity, smart grid, and digitalization.
Witold Zaklukiewicz
Data Science Team Leader at Ceneo.pl
For over six years, he has been involved in the development of artificial intelligence at Ceneo.pl, initially as a Junior Data Scientist, and currently as the Team Leader of the team. He is deeply convinced that everything happening around him can be explained by physics. After work, he enjoys cycling or locking himself in the kitchen to test new dessert recipes.
Agnieszka Niezgoda, PhD
Senior Technical Specialist, Data & AI at Microsoft
As a AI & Data Senior Technical Specialist at Microsoft, I help Microsoft customers realize the potential of data and data science. I have over a decade of experience in data & analytics, data science, and AI gained from working across industries such as Retail, FSI, Pharma and Automotive in various companies across Europe, as well as research experience from universities in Berlin and New York. I believe in combining the business and technical perspectives on IT projects to achieve the best outcomes.
Mike Guzowski
CEO, Co-Founder of Developico
Strategic IT Manager, entrepreneur and marketer. Founder of No Code Days, initiator and member of the program board of the “No Code Developer” studies at the Warsaw University of Technology, lecturer at many universities (including the Warsaw University of Technology and SWPS University). Co-founder and CEO of Developico and NoCodeMasters (Application Academy). He started his career path in IT in 2012 as a programmer, and then gained experience as a consultant, architect and team leader. In 2020, he co-founded and became CEO of two technology companies in the areas of Digital Transformation, Automation and Integration. Awarded the Microsoft MVP (Most Valuable Professional) title in the Business Applications area three times. Multiple speaker and member of program boards at leading IT conferences in the area of No Code and Low Code, co-founder of the first Power Platform community in Poland.
Łukasz Borowiecki, PhD
CEO, Co-Founder of 10 Senses
Co-founder and CEO of 10 Senses, a company providing services in the area of data science, machine learning and process mining. He holds a Phd degree in Economics from Warsaw School of Economics, and an MA degree in Sociology from Jagiellonian University in Kraków. Prior to founding 10 Senses, Łukasz worked as a consultant at EY Business Advisory (former Ernst and Young). Overall, he has over 10 years of experience in data science and data focused projects in market research and consulting companies. Łukasz is the author of several publications on the application of AI in the Polish economy published in Digital Poland Foundation.
Mikołaj Sacha
Head of Machine Learning at Molecule.one
Mikołaj Sacha works as the Head of Machine Learning in Molecule.one, a company that delivers AI-based solutions that aid chemical synthesis for drug development. He is also currently a Ph.D. candidate and a member of the GMUM research group at Jagiellonian University. Mikołaj Sacha has authored several research publications focused on the topic of interpretability in computer vision, as well as explainable AI for chemical reaction prediction and natural sciences. His other interests include AI for health care, graph neural networks, and generative models.
Marek K. Zieliński
CTO, Co-Founder of 10 Senses
Co-founder of 10 Senses, a company that offers advanced analytics products and services, including AI and process mining. Has over 10 years of experience working with data-related problems, such as data science, machine learning, artificial intelligence, and process mining. Marek specializes in Data Engineering for data visualization, ML/AI and process analytics. Has also worked on using GPUs to support computational performance. He happily shares his knowledge and experience with the market during conferences, meetings, and training sessions. He has a background in Solid State Physics and is a specialist in Data Warehousing and BI, and by hobby he is a statistical physicist.
Register