Sava Centar, Belgrade

20-24th November 2023

DSC SCHEDULE

DATA & AI TALKS

CONGRESS HALL B

Day Recap

09:15

 —

09:30

Day Recap

The Innovator’s Dilemma: to Build or Not to Build

NLP & Text Analytics
Technical talk, Research talk
Intermediate to Advanced
In this talk, we will delve into the evolving landscape of firms with established in-house machine learning teams and explore the critical question of their future role in the ever-changing world of AI. As the remit of these teams undergoes a seismic shift, a fundamental dilemma arises: Should these teams transition into AI engineering roles, leveraging commercially available tools and adapting their strategies around them, or should they continue to pioneer innovations within their specialised domains? This dilemma is unique to organizations with existing ML capabilities, setting them apart from firms without such teams. Furthermore, the talk will address the intricacies of using retrieval-augmented generation (RAG) as a potential solution in knowledge-intensive tasks, and ask whether such a method is a “hack” or a grounded methodology that will withstand the test of time. We will also reflect on the safe and responsible development of LLM-powered tools.
NLP & Text Analytics
09:30

 —

10:00

The Innovator’s Dilemma: to Build or Not to Build

Danijela Horak
Head of AI Research @ BBC

Impact of NLP in AI-powered Bank

NLP & Text Analytics
Technical talk, Business talk, Research talk
Intermediate
At NLB, we prioritize not only client satisfaction but also the experience of our employees. To that end, we have developed a sentiment analysis model that predicts the sender’s sentiment based on email content and structure. Additionally, we have a topic classification model that identifies whether clients are writing about their cards, accounts, mobile applications, loans, and more. These models help our Contact Centre employees prioritize which emails to address first, allocate resources more efficiently, and provide high-quality client service.
NLP & Text Analytics
10:00

 —

10:30

Impact of NLP in AI-powered Bank

Branislava Sandrih Todorovic & Mirella Miskic
Data Science Team Lead @ NLB DigIT, Belgrade & Executive Director of Contact center 24/7 @ NLB

Coffee break

10:30

 —

11:00

Coffee break

From Tables to Answers: building QA System for In-Document Searches

NLP & Text Analytics
Technical talk
Intermediate to Advanced
In my talk I will cover a task of tabular QA in application for unstructured pdf documents. I will walk you through each stage of the pipeline, from data preparation to modelling, and share valuable insights our team has gathered along the way.
NLP & Text Analytics
11:00

 —

11:30

From Tables to Answers: building QA System for In-Document Searches

Vladimir Ageev
Lead Data Scientist @ EPAM Systems

Multi-Task Learning with Intermediate Continual Learning for Industry NLP Use Cases

NLP & Text Analytics
Research talk
Intermediate
Companies face a problem of having to solve multiple NLP tasks while receiving new client requirements and facing distribution shifts. This talk’s objective is to find a multi-task learning (MTL) method, which is also suitable for continual learning (CL). The talk examines adapters and hypernetwork approach. The findings imply that both adapters and hypernetwork are a better alternative to STL for a MTL-CL setting. This allows companies to train the models faster, save them using less storage, integrate them easier with ML systems, and continually train them, all whilst preserving an on-par performance with STL.
NLP & Text Analytics
11:30

 —

12:00

Multi-Task Learning with Intermediate Continual Learning for Industry NLP Use Cases

Tin Ferkovic
NLP Researcher @ doXray

Coffee break

12:00

 —

12:15

Coffee break

TBA

Keynote
Keynote
12:15

 —

13:00

TBA

Jepson Taylor
Chief AI Strategist @ Dataiku

The iPhone moment of AI

Keynote
Keynote
13:00

 —

13:45

The iPhone moment of AI

Lars Rinnan
CEO @ Amesto NextBridge

Lunch break

13:45

 —

15:00

Lunch break

Capturing and Validating Real World Movements of Millions of People

Innovation with Data
Research talk
Intermediate
We present how Foursquare uses its owned and operated apps to drive quality and accuracy in movement data, to discriminate real from fake data, and to train algorithms for stop detection and venue attachment. These components power the Foursquare movement engine -reconstructing the daily diaries of millions of people every day from billions of GPS records ingested daily from across the world
Innovation with Data
15:00

 —

15:30

Capturing and Validating Real World Movements of Millions of People

Runxin Li
Staff Engineer @ Foursquare

Navigating the Horizons: AI in Media – Balancing Innovation, Quality, and Responsibility

Innovation with Data
Technical talk, Business talk
Intermediate
For TX Group, as a multi-faceted enterprise spanning media to advertising, it is crucial to explore the opportunities, challenges, ethical considerations, and social responsibility associated with AI in media. I will talk yout how we use AI applications, for example, to reflect our product, unlock valuable insights and foster transparency. Moreover, generative AI empowers journalists to meet the demand for personalized and local news while safeguarding our quality standards. By staying at the forefront of AI innovation, we can shape a future that harnesses the power of AI while upholding the core values that define our organization. This benefits both our organization and society at large.
Innovation with Data
15:30

 —

16:00

Navigating the Horizons: AI in Media – Balancing Innovation, Quality, and Responsibility

Dominic Herzog
Chief Data Scientist @ TX Group AG

Coffee break

16:00

 —

16:30

Coffee break

Data Radar Maps: Models for “data-informed” Growth in Product Management

Innovation with Data
Business talk, Research talk
Beginner to Intermediate
Innovation with Data
16:30

 —

17:00

Data Radar Maps: Models for “data-informed” Growth in Product Management

Joy Chatterjee
Founder @ The Joy of Data

From Algorithms to Assets: Data Science Meets Real Estate

Innovation with Data
Technical talk, Business talk
Beginner to Intermediate
In the evolving landscape of real estate, data science emerges as a transformative force. By combining data-driven insights with property markets, we unlock a new era of informed decision-making. A notable application is automated valuation, where machine learning leverages vast datasets to estimate property values accurately. This synergy empowers investors, agents, and buyers with comprehensive perspectives. As data science continues to shape the real estate domain, embracing automated valuation offers a promising gateway to precision and efficiency in property assessment and transactions.
Innovation with Data
17:00

 —

17:30

From Algorithms to Assets: Data Science Meets Real Estate

Ilija Lazarevic
Business Scientist, Founder @ DataBrut

New Extractivism

Closing Keynote
Educational talk
Intermediate to Advanced
Empowered by the digital extractivism tools of the information age, everything becomes a potential frontier for expansion and extraction. From the depth of DNA code in every single cell of the human organism to vast frontiers of human emotions, behavior and social relations, to nature as a whole—everything becomes the territory for the new extractivism. At this moment in the 21st century, we see a new form of extractivism that is well underway: one that reaches into the furthest corners of the biosphere and the deepest layers of human cognitive and affective being. Traditional colonial practices of control over critical assets, trade routes, natural resources and exploitation of human labor are still deeply embedded in the contemporary supply chains, logistics and assembly lines of digital content, products and infrastructure. Those relations define future development and distribution of wealth and power on a planetary scale.
Closing Keynote
17:30

 —

18:15

New Extractivism

Vladan Joler
Director @ Share Foundation

CONGRESS HALL A

Break Annex 1

09:15

 —

09:30

Break Annex 1

Make data central feature

Transforming industries with AI
Technical talk, Educational talk
Beginner to Intermediate
In traditional data workflows, we often encounter what are known as “task-oriented” interfaces and tools. These interfaces focus primarily on orchestrating and managing data tasks or jobs. Engineers use task-oriented tools like Airflow, Luigi, and Prefect to model dependencies between jobs or tasks. This approach has been more about managing, monitoring, and operating jobs than creating and managing data sets. However, a recent trend in the industry is shifting towards a more data-oriented approach. This means that instead of concentrating on the jobs and tasks (processes), organizations are now emphasizing the data sets (end products) themselves. With this data-oriented approach, jobs and tasks are still essential, but they are considered implementation details. This means that while they remain crucial in creating the data sets, they are no longer the main focus. The primary focus is now on enabling data engineers to develop data sets (data models or assets) instead of jobs. And ensuring that the data sets being generated are of high quality, easily accessible, and well-documented. The quick rise of tools like Mage.ai, Airbyte, and Dagster highlights this trend. The presentation will include a demonstration of various data-oriented tools, re-emphasizing the value of focusing on data rather than the tasks that manipulate it.
Transforming industries with AI
10:00

 —

10:30

Make data central feature

Antoni Ivanov
Staff Engineer @ VMware / Versatile Data Kit

Coffee break

10:30

 —

11:00

Coffee break

Data science in high energy physics

Applied Data Science
Research talk
Advanced
High energy physics experiments such as currently running Large Hadron Collider (LHC) or the future collider experiments (CEPC, CLIC, ILC, FCC), rely strongly on data science. Only from four LHC experiments the CERN Data Centre stores more than thirty petabytes of data per year, where over hundred petabytes of data are archived permanently. The collider experiments are characterized not only by the vast amount of data, but also with the necessity for the high precision measurement, unfavorable ratio of signal to background, where the tiny signals are covered by the huge pile of background events, with ratio of one per million, or less. In Higgs physics special challenge present the studies with purely hadronic final states, jets, where the lack of the sharp tagging variables lead to strenuous signal and background separation. The presentation will give the overview of the use of data science in the Higgs boson physics at future Circular electron positron collider, CEPC, China.
Applied Data Science
11:00

 —

11:30

Data science in high energy physics

Mila Pandurovic
Associate Research Professor @ Vinca Insitute of Nuclear Sciences

Networks Cocaine and Fruit Flies

Applied Data Science
Research talk
Intermediate
Did you know that fruit flies can even become addicted to cocaine? By studying how these tiny creatures behave when exposed to cocaine, researchers can gain a better understanding of the neurological and behavioral mechanisms underlying addiction. In this talk, we’ll explore how fruit flies have been used to study cocaine addiction and the new network-based measures that have been revealed through the analysis of their social interaction networks. This data-driven approach can help us better understand the genetic and environmental factors that influence both drug addiction and social behavior.
Applied Data Science
11:30

 —

12:00

Networks Cocaine and Fruit Flies

Milan Petrovic
PhD Student @ Faculty of Informatics and Digital Technologies

Coffee break

12:00

 —

12:15

Coffee break

Break Annex 1

12:15

 —

13:00

Break Annex 1

Gender bias in AI – In collaboration with Women in Data Science Belgrade

Panel Discussion
Panel Discussion
13:00

 —

13:45

Gender bias in AI – In collaboration with Women in Data Science Belgrade

Natali Delic & Isidora Gataric & Ljubica Vujovic Boskovic, Moderated by Marija Novakovic
Chief Strategy and Digital Officer (CSO, CDO) and Executive Board Member @ Telekom Srbija & Analytics Delivery Lead @ Wireless Media Group & Program Manager 2 @ Microsoft, Program Director @ DSC Franchize

Lunch break

13:45

 —

15:00

Lunch break

Under the Hood: Key Algorithms Powering Vector Search

Edge & Distributed Computing
Technical talk
Intermediate to Advanced
Join Dusan Milunovic, data engineer at SmartCat as he unravels the core algorithms driving vector search, focusing on HNSW, LSH, and PQ. This session offers a concise exploration of the pivotal role these algorithms play in the landscape of AI-driven search. Dive deeper into the world of AI-driven search, and grasp the technical intricacies that fuel it. Suitable for anyone curious about the tech behind vector databases.
Edge & Distributed Computing
15:00

 —

15:30

Under the Hood: Key Algorithms Powering Vector Search

Dusan Milunovic
Data Engineer @ Smartcat

You will need custom chips soon: where high-performance and energy-efficent computing is going

Edge & Distributed Computing
Technical talk
Intermediate to Advanced
You’re unlikely to see another 2x improvement in single-threaded performance, so where must you go for more compute? Manycore & distributed computing? Custom hardware accelerators? Reconfigurable & adaptive systems? In this talk I will address the concerns of those on the latency-sensitive and energy-bound side of computing, and will show a path to building custom silicon without in-house hardware experts.
Edge & Distributed Computing
15:30

 —

16:00

You will need custom chips soon: where high-performance and energy-efficent computing is going

Mihailo Isakov
Founder & CEO @ BoolSi

Coffee break

16:00

 —

16:30

Coffee break

Applied Custom Vision

Edge & Distributed Computing
Technical talk
Beginner to Intermediate
Al is the buzz, chatGPT, CoPilots everywhere. So AI ca do anything now, no?. And are a lot of camera feeds already available at our workplaces from security to QA. How to really “use” those feeds? How to get information (are 12 cars in the parking), alerts (roses heed water), make a difference (5 peoples are in the crane danger radius). This session will show what we can do, relatively easily, with a live video feed, using Custom Vision on the edge. Demos end to end, KubeAI Application Nucleus for edge, 5G, OpenVino
Edge & Distributed Computing
16:30

 —

17:00

Applied Custom Vision

Catalin Gheorghiu
Solution Architect @ EPAM Systems

CONGRESS HALL C

Break annex 2

09:15

 —

09:30

Break annex 2

Approach to Architecting Generative AI Solutions

Deploying ML Models
Technical talk
Intermediate to Advanced
Exploring how complex AI agents are built and integrated in solutions from the solutioning perspective. Share insights from architecting, designing, and developing solutions based on Generative AI.
Deploying ML Models
09:30

 —

10:00

Approach to Architecting Generative AI Solutions

Ivan Petrovic
Senior Machine Learning Technology Lead @ HTEC

Under Pressure: Applying ML in Real Time

Deploying ML Models
Technical talk
Intermediate
In today’s world, more and more industries increasingly rely on ML to improve their business, and a large number of them require near-instant processing in a streaming scenario – from autonomous vehicles and healthcare monitoring, all the way to fraud detection and recommendation systems. How can we create systems that work with extremely low latencies? How can we make these systems scalable? How to ensure high availability without data loss? This talk attempts to answer these questions and more.
Deploying ML Models
10:00

 —

10:30

Under Pressure: Applying ML in Real Time

Miroslav Bicanic
ML Engineer @ CROZ AI d.o.o.

Coffee break

10:30

 —

11:00

Coffee break

Deploying computer vision models at scale devops-free

Deploying ML Models
Technical talk
Intermediate
This talk will explain how we deploy computer vision models at Veriff by leveraging a mix of classic software engineering techniques combined with last AWS tooling. Thanks to this, we have cut down our model deployment time from 10 days to 1 day and inference costs by 75%
Deploying ML Models
11:00

 —

11:30

Deploying computer vision models at scale devops-free

Ricard Borras Navarra
Senior Machine Learning Engineer @ Veriff

“Unit tests for Data” with Deequ

Deploying ML Models
Technical talk
Intermediate
Modern companies and institutions rely on data to guide every single business process and decision. Missing or incorrect information seriously compromises any decision process downstream. Its 2023, everyone knows that its not about tuning models its about volume and cleanliness of data. Your customers should not be your “framework” for testing the data in production. Deequ is a system that offers declarative API that combines standard quality constraints wutih user-defined validation code allowing “unit tests” on data just as in software engineering realm. Its efficiently executes constraint validation using Apache Spark, supporting incremental validation on growing datasets.
Deploying ML Models
11:30

 —

12:00

“Unit tests for Data” with Deequ

Pavle Tabandzelic
MLOps Engineer @ Raiffeisen Bank

Coffee break

12:00

 —

12:15

Coffee break

Break annex 2

12:15

 —

13:00

Break annex 2

Can AI transform business?

Panel Discussion
Panel Discussion
13:00

 —

13:45

Can AI transform business?

Tobias Zwingmann & Venkat Sivasubramanian & Alin Kalam, Moderated by Martina Silov
Author & Managing Partner @ RAPYD.AI & Senior Director, IoT, Data, Analytics, ML @ Rivian & Head of Market Intelligence & Retail/Digital Data Strategy @ UNIQA, Executive Director @ CroAI

Lunch break

13:45

 —

15:00

Lunch break

Brain-Computer Interfaces: Ethical issues and resolutions

Ethical AI & AI4Good
Business talk
Beginner to Intermediate
Brain-Computer Interfaces (BCI) are a type of direct communication between the human brain and its functions and a computer that allow the human to control an external object, such as an artificial limb, or allow a digital device to control the human brain—for instance, to detect and stop epileptic seizures. A new form of BCI combined with AI is already being used experimentally for emotional therapy in psychiatric patients. I will outline this case and then discuss the ethical problems and possible resolutions that a corporation that makes AI-based appliances might consider in order to enhance the device’s ethicality, marketability, and the company’s reputation.
Ethical AI & AI4Good
15:00

 —

15:30

Brain-Computer Interfaces: Ethical issues and resolutions

Kevin LaGrandeur
Fellow, and Professor @ Institute for Ethics & Emerging Technology, and NYIT

Responsible AI 2.0: NO AI without RAI

Ethical AI & AI4Good
Educational talk
Intermediate
In the last 10 years machine learning has become ubiquitous and touches all lives in ways that was unimaginable before. The machines can make decisions that required considerable human effort at a much faster speed and reduced cost with a little human oversight. As a result, machines don’t just have a higher than before influence in shaping our lives but are also under increased scrutiny by both regulators as well as user rights advocates. The adage “with great power comes great responsibility” has long been used – from French revolution to superhero comics. It has never been truer as the great power that machine learning wields is now in the hands of almost anyone making a software product. It ranges from giving people access to the funds that can alter their lifepath, medical diagnosis that can increase their life expectancy or reduce it dramatically to their social media feed that cannot just provide them the content that keeps them engaged, but also polarise their beliefs by feeding them information that reinforces their existing notions. With the growing influence of AI technologies and the corresponding scrutiny, the way AI development happens is beginning to change. The full data science lifecycle needs to incorporate the elements of responsible AI viz. Fairness, accountability, privacy & transparency and the professionals who know how to design and implement these will be the ones that employers will look for. Not to ignore the issue of hallucination, monetisation, robustness, reliability and attribution that needs to be address as larger Responsible AI framework. Along with this, the RAI principle needs to consider human centricity, planet inclusive, dsaftety and societal well-being.
Ethical AI & AI4Good
15:30

 —

16:00

Responsible AI 2.0: NO AI without RAI

Sray Agarwal
Director Responsible AI @ Fractal

Coffee break

16:00

 —

16:30

Coffee break

AI and Geopolitical Strategy

Ethical AI & AI4Good
Research talk, Educational talk
Intermediate to Advanced
A new era of geopolitical competition will be marked by advanced technological development. Great Powers are adopting complex strategies trying to shift the balance of power. AI, as a cornerstone of the strategy for the new century, reflects both the strategic vision and the societal design of China, Russia, the EU, and the United States. What new forms of risks spurring from AI adoption and how Great Powers’ competition may look, remains to be seen. This work attempts to provide some insights related to a new cyber-geopolitical landscape.
Ethical AI & AI4Good
16:30

 —

17:00

AI and Geopolitical Strategy

Dragan Vitorovic
HFRI Scholar and Researcher @ Aristotle University of Thessaloniki

Ethical Considerations in Predictive Analytics

Ethical AI & AI4Good
Business talk, Research talk
Beginner to Intermediate
As the data-driven landscape rapidly evolves, predictive analytics holds tremendous potential for transformative insights, with predictive models becoming integral to decision-making. However, this immense power demands an equally profound responsibility towards ethical considerations. In this talk, we delve into the crucial interplay between predictive analytics and three paramount ethical aspects: data privacy, bias mitigation, and accountability. We will explore strategies for safeguarding sensitive information, mitigating bias in algorithmic decision-making, and fostering transparency to ensure accountability. Join us to delve into the ethical dimensions of predictive analytics.
Ethical AI & AI4Good
17:00

 —

17:30

Ethical Considerations in Predictive Analytics

Bunmi Akinremi
Microsoft Certified AI Engineer, Machine Learning Engineer @ Kochava

MR 7

Break DSC:X

09:15

 —

09:30

Break DSC:X

When Anxiety meets Decision making – a (personal) story on how Therapy can help

AI & Data Product Development
Transformational talk
Beginner to Intermediate
I’ve found myself in the decision making chair just enough times in order develop a love-hate relationship with it. It caused me to celebrate, it caused me to suffer, but most importantly, it forced me into a burnout and 7+ years of REBT therapy. This session will uncover how Therapy can tremendously help in Decision Making and Execution process.
AI & Data Product Development
09:30

 —

10:00

When Anxiety meets Decision making – a (personal) story on how Therapy can help

Mihailo Joksimovic
Senior Software Engineer, Microsoft

Why Governments Should Build Their Own AI Products

AI & Data Product Development
In this presentation I juxtapose (well known) arguments why governments should stay away from IT development, with (much less well known, perhaps emerging) arguments why they in fact should delve into AI development. The presentation is a blend of theoretical (IT and data product) management literature, with our own hands-on experience managing AI implementations.
AI & Data Product Development
10:00

 —

10:30

Why Governments Should Build Their Own AI Products

Vid Stimac
Project Manager, UNDP Serbia

Coffee break

10:30

 —

11:00

Coffee break

How to treat your data as a product

AI & Data Product Development
Technical talk, Business talk
Beginner to Intermediate
In today’s digital age, data has become a valuable asset that can drive innovation, improve decision-making, and enhance business performance. However, many organizations still struggle to fully leverage the potential of their data. This presentation aims to shed light on the concept of treating data as a product and how it can transform the way we handle and derive value from data. We highlight the importance of data as a strategic asset and discuss key principles such as data governance, quality management, data contract and data monetization.
AI & Data Product Development
11:00

 —

11:30

How to treat your data as a product

Ivan Dundovic
Data Architect, CROZ

How can we use AI to bring new value to the business

AI & Data Product Development
Use case, Transformational talk
Intermediate
SmartCat is a data company that works with other companies to help them build AI solutions. We are a blend of data scientists and data engineers and that makes us question from different angles how the next big AI module will be integrated in your platform. This is the prime reason why we can brag that we have more than dozen AI solutions in production developed over the last 8 years. In this talk we will share our experience working with various clients on AI solutions. We will give you a list of the 10 most common AI pitfalls that prevent AI solutions from ending in production. Are you familiar with PoC drawers, where RnD departments allocate some money to try something new, build that up to a working solution, but it never ends up in production? This list will help you prepare for your next AI project and it will help you lower down the chance for it to end up in the PoC drawer.
AI & Data Product Development
11:30

 —

12:00

How can we use AI to bring new value to the business

Nenad Bozic
CoFounder & CEO, SmartCat

Coffee break

12:00

 —

12:15

Coffee break

Cloud Advisory: Cloudy, with a chance of people

Keynote
Keynote
12:15

 —

13:00

Cloud Advisory: Cloudy, with a chance of people

Josema Cavanillas
Global Head of HCI Technology Consulting @ Atos

Ethics by Design: The Role and Responsibility of AI Engineers in Shaping Responsible Generative AI

Keynote
Transformational talk
Intermediate
As generative AI grows more advanced and pervasive, AI engineers find themselves in positions of unseen power and influence. With this power comes great responsibility: a duty to consider the ethical implications of their work at every stage of the development process. In this talk, I provide practical examples of how engineers can embed ethical considerations into their workflows and collaborate with ethics experts. My aim is to provide a glimpse into AI development and provide actionable ways engineers can develop generative AI applications responsibly.
Keynote
13:00

 —

13:45

Ethics by Design: The Role and Responsibility of AI Engineers in Shaping Responsible Generative AI

Heiko Hotz
Founder, NLP London

Lunch Break

13:45

 —

15:00

Lunch Break

Building successful data teams in the realities of 2023

Building Data & AI Teams
Transformational talk
Beginner to Intermediate
In this talk I will address: Why and how to address business value added by the data team? What to do in your first 2 months as Head of Data or CDO? Balancing between business problems solved fast & slow. Dealing with technical debt, legacy teams, and legacy tech. When to set up central or decentral data teams? The ROI of the Data team and the risk cost of success. Building the skills and competencies to run a Data A-team
Building Data & AI Teams
15:00

 —

15:30

Building successful data teams in the realities of 2023

Karl Ivo Sokolov
Managing Partner | Data, Specific-Group Austria

Building Bridges: Connecting Business and Tech for Successful Data Science Projects

Building Data & AI Teams
Business talk
Beginner to Intermediate
If you’re a business or tech professional, you’ve probably had a data science project fall apart in your hands. In this talk, we discuss challenges teams face when working together and how they could overcome them if they understood, what each side needs to be successful. We talk about skills that are needed in a project team to ensure that projects go smoothly into production – and stay there. With examples from practice, we hope to inspire you to approach data science projects with a flexible, collaborative mindset that emphasizes communication, diversity of skill sets, and clear goals and metrics.
Building Data & AI Teams
15:30

 —

16:00

Building Bridges: Connecting Business and Tech for Successful Data Science Projects

Nina Mrzelj
Senior Data Scientist, Sclable Business Solutions GmbH

Coffee Break

16:00

 —

16:30

Coffee Break

Remote work is here to stay – and what’s next? Asynchronous way – the next level of remote work

Building Data & AI Teams
Business talk, Educational talk
Beginner to Intermediate
A great chance to hear the experience about the future of work from the biggest all-remote company in the world. Arm yourself with the best-practices toolbox for the async work tips and tricks and leverage the context from the distributed work leaders.From the GitLab Data Team member, a first-face story about how the asynchronous way of working helps us manage the “chaos” (as folks usually think about remote work) in the 24/7 work environment. Vibrant session to expose the picture of the pioneers and advocates of all-remote philosophy. Forget working hours, timesheets, offices, and emails and join the ride. Will show you how this way of work brings joy to me and my 1600 colleagues from 56 countries and how we serve 30+ million customers with exactly 0 offices. – Is remote work just a perk? – Various approaches to remote strategies – Why “just” remote work is not enough? – What’s the next level for remote work?
Building Data & AI Teams
16:30

 —

17:00

Remote work is here to stay – and what’s next? Asynchronous way – the next level of remote work

Radovan Bacovic
Staff Data Engineer, GitLab

MR 11

ChatBot Learning Assistant with Large Language Models

Transforming industries with AI
Technical talk
Intermediate to Advanced
Join this talk as I reveal the fascinating journey of creating a cutting-edge bot assistant powered by Large Language Models, paving the way from POC to MVP. Prepare to dive deeper into: A unique multi-route architecture that forms the backbone of our bot’s performance., Ingenious solutions to combat latency challenges for a seamless user experience., Techniques to tackle out-of-distribution questions and enhance the bot’s adaptability. Don’t miss the opportunity to discover valuable insights that could elevate your own bot development process. See you there!
Transforming industries with AI
09:30

 —

10:00

ChatBot Learning Assistant with Large Language Models

Vladislav Belov
Data Scientist @ EPAM Systems

Transforming Business Process Automation with Retrieval-Augmented Generation and LLMs

Transforming industries with AI
Technical talk, Business talk
Intermediate
Transforming Business Process Automation with Retrieval-Augmented Generation and LLMs
Transforming industries with AI
10:00

 —

10:30

Transforming Business Process Automation with Retrieval-Augmented Generation and LLMs

Djordje Grozdic
Senior Staff Data Scientist @ Grid Dynamics

Coffee break

10:30

 —

11:00

Coffee break

A Hybrid LLM based Enterprise Chatbot Solution Weaver Peer by Intellya

Transforming industries with AI
Technical talk
Intermediate to Advanced
In the rapidly evolving landscape of artificial intelligence and customer engagement, Intellya presents its innovative product, Weaver Peer. This advanced solution seamlessly integrates domain-based chatbots and Large Language Model (LLM) chatbots, offering a versatile and efficient approach to handling customer queries.
Transforming industries with AI
11:00

 —

11:30

A Hybrid LLM based Enterprise Chatbot Solution Weaver Peer by Intellya

Andrija Petrovic
Senior Machine Learning Architect @ Intellya

Innovating the Telecom Industry: Embracing Automation and AI

Transforming industries with AI
Business talk
Intermediate
“My speech explores the transformative impact of automation and artificial intelligence (AI) within the telecommunications sector. Commencing with an overview of the telecom industry’s landscape and the pivotal role of process automation and AI, the presentation delves into defining these concepts and their significance. It emphasizes the role of AI in fostering innovation and streamlining operational efficiency. Furthermore, my speech highlights the potential applications of automation and AI in improving customer experience, ranging from personalized services, introduction of new products and enhancing overall network performance. The benefits stemming from their implementation, such as increased productivity, cost reduction, exploring new markets and advanced predictive analytics, are discussed. The strategies for successful integration, including infrastructure development, workforce upskilling, collaborations, and ethical considerations, are outlined to guide the audience through effective implementation. Additionally, potential challenges such as data privacy, resistance to change, and regulatory obstacles are addressed, paving the way for a comprehensive understanding. Concluding with key insights, future trends, and a call to action for embracing these technologies, the speech encourages a proactive approach toward the inevitable transformations in the telecom industry. The session will be followed by a Q&A session, fostering further discourse, and a strong appeal to embrace the possibilities presented by automation and AI.”
Transforming industries with AI
11:30

 —

12:00

Innovating the Telecom Industry: Embracing Automation and AI

Nebojsa Dobrijevic
COO @ Orion Telekom

Coffee break

16:00

 —

16:30

Coffee break

BREAKOUT SESSIONS

MR 8

Intelligent Farming

09:30 – 09:45   Blending IoT and AI – experiences from the “kitchen”

Srdjan Krco
CEO/founder @ DunavNET

In DunavNET’s “kitchen”, a variety of specialties have been crafted for years in agricultural and manufacturing domains, using digital technologies as essential ingredients. This presentation focuses on sharing experiences in the fusion of IoT and AI technologies within the context of Intelligent Farming. The discussion will delve into how the synergy of these technologies transforms farming practices, optimizing operations, and enhancing sustainability. Join the session for a journey that reveals the potential of IoT and AI in reshaping the future of agriculture.



09:45 – 10:00   AI in Agriculture

Oskar Marko
Assistant Director @ BioSense Institute

Oskar’s talk will focus on the AI research conducted at BioSense Institute, the European Centre of Excellence for AgTech. He will present the state of agriculture today and explain why we need technology more than ever, to ensure both profitability and sustainability of the production. Through a number of use cases, such as agricultural robotics, predictive models, and decision-support systems, we will see how AI is transforming agriculture and how BioSense’s systems are used on more than 1/4 of the whole Serbian cropland.



10:00 – 10:15   Developing AI-powered hardware for crop farming

Filip Injac
Co-Founder @ Atar AI

How we have developed ATAR smart sprayer, an AI-powered system that reduces chemicals usage in fields up to 80% by enabling any existing spraying machine to perform precise weed detection and spot spraying.



10:15 – 10:30   Industry 4.0: A New Hope in Farming Through Data

Jelena Zoric
Data Analyst @ Cropt

Environmental footprint and food security have become some of the most important topics in recent years. Naturally, farming is a field which greatly affects both areas. So, what can we do to minimize those risks while also giving the farmer the best conditions for success? This is just one of the questions Cropt is trying to answer by predicting both yields and risks, and finding the optimal sowing solutions. Of course, the story doesn’t end here, so we will also discuss how such data can be applied to other agricultural challenges, such as securing access to finance and optimising logistics.



10:30 – 10:45   AI in beekeeping – can we understand bees?

Jelena Pejic
CTO @ Beehold

Imagine that you can watch bees making honey from your home! And that you receive notification when it is time for you to collect it! Beehold – we develop AI assistant that allows beekeepers to inspect beehives in real time and get advice for optimal beekeeping 24/7. For the first time in our history, we have developed hardware that allows visual inspection of the beehive without opening it . The collected data is enriched with indirect information from other IoT devices, making Beehold AI Assistant a powerful, hardware-agnostic tool that can improve beekeeping process and human relationship with honeybees all over the world.



10:45 – 11:30   Panel discussion

Srdjan Krco & Oskar Marko & Filip Injac & Jelena Zoric & Jelena Pejic
CEO/founder @ DunavNET & Assistant Director @ BioSense Institute & Co-Founder @ Atar AI & Data Analyst @ Cropt & CTO @ Beehold
09:30

 —

11:30

Intelligent Farming

Srdjan Krco, Oskar Marko, Filip Injac, Jelena Zoric, Jelena Pejic

Prompt Engineering

14:30 – 14:45   Don’t be a Prompt Engineer, be a Smart Engineer

Nikola Jankovic
Backend Engineer @ SmartCat.io

With the rise of AI tools and the usage of them from gardeners to programmers, we will try and take a step back and discuss what should mean “I did this with the help of ChatGPT”. This panel will try and showcase how and when we should use these tools to help us boost our efficiency and productivity as a day to day software engineer.



14:45 – 15:00   Lessons Learned: Enhancing Visual Storytelling with MidJourney AI in Comics

Darko Dimitrijevic
UX/UI Graphic Designer @ PlayStudios

Leveraging the capabilities of MidJourney and ChatGPT while delving into the realm of comics and storytelling. The process involves the meticulous task of ensuring character consistency through MidJourney with prompt and parameter modifications, while simultaneously depicting characters in various poses, gestures, and diverse facial expressions. Extensive use of Photoshop is integral in weaving these separate visual elements into a coherent and singular composition. Additionally, ChatGPT plays a pivotal role in the conversion of text from traditional storytelling to the dynamic format of comics. This showcase is an exploration of the dynamic interplay between AI and creativity, transforming stories into visually captivating comic narratives.



15:00 – 15:15   Unlocking User Intent: The Art of Crafting Effective Prompts

Milan Roncevic
Business Process Automation Specialist @ RonTech Agency

Understanding user intent is crucial in the field of Artificial Intelligence. Prompt engineering is key to bridging the communication between users and AI models. This presentation will address common challenges in prompt design such as ambiguity and bias, and will introduce effective techniques for crafting prompts. Through real-world examples, I will demonstrate the significant impact of well-designed prompts in areas like customer support and content generation. Ethical considerations in prompt engineering and useful tools and resources will also be discussed. This presentation aims to provide attendees with insights and practical approaches to better interpret user intent through refined prompt engineering.



15:15 – 15:45   Panel discussion

Goran S Milovanovic & Nikola Jankovic & Darko Dimitrijevic & Milan Roncevic
Lead Data Scientist | Owner @ Smartocto | Data Kolektiv & Backend Engineer @ SmartCat.io & UX/UI Graphic Designer @ PlayStudios & Business Process Automation Specialist @ RonTech Agency
14:30

 —

16:00

Prompt Engineering

Goran S Milovanovic, Nikola Jankovic, Darko Dimitrijevic, Milan Roncevic

MR 10

Use of AI in Marketing

09:30 – 9:45   AI-Powered Marketing: Unleashing the Kraken

Michael Sigamani
Chief Data Officer @ Qudo AI

Unlocking the full potential of AI, we proudly showcase a game-changing paradigm shift in marketing strategy. By tapping into the goldmine of zero-party survey data, and the wealth of machine learning tools at our fingertips, we empower ourselves to transcend traditional assumptions and directly engage with customers’ preferences, needs, and aspirations. This personalised approach eliminates guesswork, establishing a direct bridge between our offerings and consumers’ desires. Our presentation promises an insightful exploration of how the convergence of zero-party survey data and self-correcting feedback loops lays the groundwork for a new era of analytical marketing. We invite you to be part of this transformative journey and witness firsthand how our innovative fusion of AI and marketing redefines success.



09:45 – 10:00   AI & HI for better marketing strategy

Marijana Agic-Molnar
CEO @ Data Do

Out of all industries, marketing industry has been and will be the highest user of AI. Marketing efficiency has significantly increased over time (with recent strong push of generative AI) especially in marketing operations – but what about marketing strategy and management? In front of marketing decision makers there are important questions to be answered: how AI may support strategic thinking and decision making? Which steps may/ can/ should stay under a human supervision and which strategic decisions may be “outsourced” to AI? How to orchestrate complexity of tools and enhance HI & AI synergy? These questions are still open for marketing practitioners and need to be addressed in line with impressive improvements in AI marketing tools.



10:00 – 10:15   Self-service BI – The myth and the reality (lessons for AI tool creators)

Fatima Soomro & Luca Pescatore
Chief Operations Officer @ Argusa SA, Switzerland & Head of Technology @ Argusa SA, Switzerland

We will explore how the humble analyses using spread sheets have evolved into curated sets “data assets”. The expectations, the process of creating a self-service BI and the current status. We will try to draw some lessons for the next generation of data tools.



10:15 – 10:30   Users behavior dynamics — modeling and predictions

Milan Bradonjic
Senior Software Engineer @ Schneider Electric

In this talk we present how to examine users’ behaviors in (very) large complex systems over time. We ask if there are certain patterns that users’ examine, and if so how to model and to identify those patterns. We present how one can develop, and provide a distributed computing solution. Moreover, we relate the above to some of global optimization problems in large graphs and data sets.



10:30 – 10:45   AI Usecases in Marketing

Luka Anicin
Founder & AI/Data Consultant @ Datablooz



10:45 – 11:30   Panel discussion

Marijana Agic-Molnar & Fatima Soomro & Luca Pescatore & Milan Bradonjic & Luka Anicin
CEO @ Data Do & Chief Operations Officer @ Argusa SA, Switzerland & Head of Technology @ Argusa SA, Switzerland & Senior Software Engineer @ Schneider Electric & Founder, AI/Data Consultant @ Datablooz
09:30

 —

11:30

Use of AI in Marketing

Marijana Agic-Molnar, Michael Sigamani, Fatima Soomro, Luca Pescatore, Milan Bradonjic, Luka Anicin

Smart Energy & AI

14:30 – 14:50   Wired for Intelligence: Unleashing AI in Power Grids Management

Vladan Krsman
Vice President – Grid operations management @ Schneider Electric Hub



14:50 – 15:10   Deep Learning models for multivariate heat demand forecasting

Milan Zdravkovic
Associate Professor @ Faculty of Mechanical Engineering University of Nis



15:10 – 15:30   Data-Driven Framework for Sustainable Manufacturing

Bratislav Trojic
Data Scientist @ Nissatech Innovation Centre

Join us on a transformative journey where we enable small and medium-sized manufacturing companies to harness the power of data, improve their processes and reduce energy waste and carbon footprint. Our innovative and comprehensive system integrates data collection through sensor installation, real-time data observability for quality control, and advanced data analysis featuring machine and deep learning techniques to gain insights into process behaviors, detecting anomalies, process variances and even unusualities in the operating of the manufacturing system. We are committed to enhancing manufacturing efficiency and sustainability through smart energy-driven optimization.



15:30 – 16:00   Panel discussion

Vladan Krsman & Milan Zdravkovic & Bratislav Trojic
Vice President – Grid operations management @ Schneider Electric Hub & Associate Professor @ Faculty of Mechanical Engineering, University of Nis & Data Scientist @ Nissatech Innovation Centre
14:30

 —

16:00

Smart Energy & AI

Vladan Krsman, Milan Zdravkovic, Bratislav Trojic

MR 12

Augmented Analytics in sport

09:30 – 9:45  Match Fixing Detection

Stefano Gemin & Domen Pozrl
Data Science Team Lead @ Sportradar & Data Scientist @ Sportradar

The Sportradar Universal Fraud Detection System monitors over 900,000 sport events per year, reporting suspicious betting activities to hundreds of sport federations across 30+ sports. To do this, over 5 billions betting transactions and 25 billions odds movements from 600+ betting operators are processed each year in real time by a ML model, with suspicious events being flagged for removal and integrity analysts leveraging prediction explanations for further analysis.



09:45 – 10:00   Injury prediction in Soccer: Will it ever happen?

Arian Skoki
Research & Teaching Assistant @ Faculty of Engineering, University of Rijeka

Soccer, a dynamic sport combining skill and chance, poses a unique challenge for injury prediction. As technology advances, the prospect of leveraging machine learning to anticipate injuries gains momentum. However, the unpredictable nature of the game, coupled with individual variability and the complexity of injury mechanisms, raises questions about the feasibility of accurate predictions. This talk delves into the complexities of soccer injury dynamics, explores the current state of machine learning applications, and examines the hurdles hindering reliable injury forecasts, and confronts the problems associated with acquiring and preparing comprehensive datasets.



10:00 – 10:15   Data-Driven Player Development: Nurturing Future Champions – Today

Marko Marjanovic
Chief Digital and Innovation Officer @ Artifit



Player development is a highly complex activity with numerous challenges, limited time for action and no room for error. In the quest for new perspectives, numerous scientific disciplines are involved, leading to the creation of enormous data sets that need to be collected, aggregated, and analyzed to serve as foundation for managing the process of selecting and training young footballers. How might data insights support coaches in developing future champions?

10:15 – 10:30   Data-driven Match Analysis in Soccer

Henrik Biermann
Data Analyst @ Borussia Dortmund



While most data-driven analytics in soccer is still performed with qualitative video data, there exist two sources of soccer that recently gained popularity. Event data, comprising the player’s actions on the field in sport-specific terms, and position data, describing the movement of the players and the ball in two coordinates. As simple integration of these data sources can already lead to an improvement of usual workflows in soccer analytics (e.g., automatic retrieval of certain situations via event data), more complex (automatic) approaches are already developed and aim to offer a more substantial extension of the field up to the generation of new, unknown insights. In this talk, we discuss how data can improve the workflow of professional football analytics and shed light on the barriers that currently prevent the integration of more sophisticated AI algorithms in the sport.

10:30 – 11:00   Panel discussion

Stefano Gemin & Domen Pozrl & Arian Skoki & Marko Marjanovic & Henrik Biermann
Data Science Team Lead @ Sportradar & Data Scientist @ Sportradar & Research & Teaching Assistant @ Faculty of Engineering, University of Rijeka & Chief Digital and Innovation Officer @ Artifit & Data Analyst @ Borussia Dortmund
09:30

 —

11:00

Augmented Analytics in sport

Stefano Gemin, Domen Pozrl, Arian Skoki, Marko Marjanovic, Henrik Biermann

Breakout Break BS3

11:00

 —

11:30

Breakout Break BS3

NeuroAI

15:00 – 15:15   Cyberpunk revolution

Max Talanov
Senior research associate @ Institute for artificial intelligence

The talk is dedicated to the set of technologies emerging at the moment including BCI, AI and neuromorphic technology that now are forming the foundation for a new generation breaking through approaches and applications. We see the paradigm shift in the understanding of how technology could influence and merge with humans and thus what is the human body. I’ll provide a vision of the future of the AI + humans perspectives that we call the Cyberpunk revolution.



15:15 – 15:30   Fungal intelligence

Andrew Adamatzky
Professor @ UWE Bristol

The fungi is the largest, widely distributed and the oldest group of living organisms. The smallest fungi are microscopic single cells. The largest mycelium occupies over dozens of hectares and weighs several tons. Fungi possess almost all the senses as humans do. They sense light, chemicals, gases, gravity and electric fields. Fungi exhibit an electrical response to stimulation in a matter of seconds or minutes. We demonstrate that fungi process information in a manner similar to neurons. We represent fungi from an unusual perspective—as sensors, electronic neuromorphic devices and future computers. We propose fungal electronics, computing and intelligence as an alternative to currently adopted methodologies.



15:30 – 15:45   Neuromorphic systems based on organic memristive devices

Victor Erokhin
Research Director @ Institute of Materials for Electronics and Magnetism, National Research Council (IMEM-CNR)

There are special requirements for the hardware realization of neuromorphic systems, namely:
Same elements must be used for memorizing and processing of the information;
It must allow parallel information processing;
It must have internal activity (cross-talk of elements, oscillation processes, effect of noise, etc.)
It must allow bio-realistic learning algorithms;
It must allow the formation of 3D systems with distributed connection of elements;
It must allow interfacing with nervous system.

We will consider the importance of each item and approaches for their practical realization.



15:45 – 16:00   Rewiring Minds and Bodies

Jordi Vallverdú
Professor @ ICREA – UAB

Join us to explore the transformative impact of AI on Human-Robot Interaction (HRI) and Man-Machine Interfaces, with a focus on the intriguing concept of ‘rewiring bodies.’ Discover how AI-driven systems enhance interactions, empower robots to adapt to human preferences, and enable individuals to control machines through brain-computer interfaces and wearables. Dive into the forefront of HRI and interface technology, where AI is revolutionizing the way we understand, collaborate, and augment our physical and cognitive capabilities.



16:00 – 16:15   NeuroAI Hardware Accelerators

Rastislav Struharik
Professor @ Faculty of Technical Sciences, University of Novi Sad

Current state-of-the-art AI systems, based on 2nd generation neural networks (Convolutional Neural Networks, Large Language Models, etc.), have proved to be highly useful in many application domains, however, their performance in terms of computational efficiency (speed and power consumption) makes them unsuitable for edge-based systems. The 3rd generation of neural networks, Spiking Neural Networks (SNNs), have been proposed to bridge the gap between the biological and artificial neural computing systems, using biologically plausible models of neurons to fully exploit the efficiency in the spatiotemporal signal coding and processing, and the corresponding power efficiency observed in the BNSs. Spiking neurons operate asynchronously, each neuron firing an output spike only when its state reaches a certain threshold. SNN information processing is event-driven, which makes a crucial difference between the 2nd and 3rd generation of ANNs. Although highly promising, current SNN technology has not reached maturity yet, lagging behind state-of-the-art 2nd generation ANN systems in classification accuracy. One way of moving forward is to develop more efficient SNN training algorithms, use more advanced spike information encoding schemes, etc. Another alternative, offering a mid-term solution, is to develop 2nd–3rd gen hybrid ANN/SNN systems, combining the advantages of both approaches. These hybrid systems could offer a comparable classification performance of current 2nd gen ANNs, while still being significantly more power efficient, due to the event-based processing, inherent to the 3rd gen SNN architectures. This talk will present existing results and future work on the development of hardware accelerators for 2nd and 3rd-generation ANN systems that is being conducted at the University of Novi Sad.



16:15 – 16:30   Memristive Spiking Neural Networks

Natasa Samardžic
Associate Professor @ Faculty of Technical Sciences, University of Novi Sad

Neuromorphic computing, a computing approach directly inspired by human brain, shows significant potential in addressing the von Neuman bottleneck. Spiking neural networks (SNNs) represent a category of artificial neural networks which model behavior of biological neurons and synaptic connections are widely used in the domain of neuromorphic systems. Memristors (memory-resistors), as basic two-terminal circuit components offer unique properties including nonlinearity, multilevel resistance state, energy efficiency and nanoscale size. We will discuss the potential and opportunities of the implementation of memristors in spiking neural networks providing both synaptic plasticity and neuron spiking behavior.



16:30 – 17:00   Panel discussion

Max Talanov & Andrew Adamatzky & Victor Erokhin & Jordi Vallverdu & Rastislav Struharik & Natasa Samardzic
Senior research associate @ Institute for artificial intelligence & Professor @ UWE Bristol & Research Director @ Institute of Materials for Electronics and Magnetism, National Research Council (IMEM-CNR) & Professor @ ICREA – UAB & Professor @ Faculty of Technical Sciences, University of Novi Sad & Associate Professor @ Faculty of Technical Sciences, University of Novi Sad
15:00

 —

17:00

NeuroAI

Max Talanov, Andrew Adamatzky, Victor Erokhin, Jordi Vallverdu, Rastislav Struharik, Natasa Samardzic

CO-LOCATED EVENTS

MR 9

AI Impact on Cybersecurity and Privacy

Keynote
Keynote
09:45

 —

10:30

AI Impact on Cybersecurity and Privacy

Dragan Pleskonjic, Vladimir Jelic
Senior Director, Manager Application Security, IGT

New age IT – Cybersecurity challanges with AI

Security-focused AI
Business talk
Intermediate
cyber security and it´s challenges in AI
Security-focused AI
10:00

 —

10:30

New age IT – Cybersecurity challanges with AI

Amra Drnda
Applied Data Science, ISG