04 - 05 October 2022

Deep Learning Summit Deep Learning Summit schedule

Berlin AI Summit



Download PDF
  • 08:00

    REGISTRATION & LIGHT BREAKFAST

  • 09:00

    WELCOME NOTE & OPENING REMARKS

  • LATEST ADVANCEMENTS IN DL

  • 09:15
    Srayanta Mukherjee

    Representation Learning on Graphs: Applications & Innvovations Across the Pharma Value Chain

    Srayanta Mukherjee - Director - Data Science & AI - Novartis

    Down arrow blue

    Representation Learning on Graphs: Applications and Innovations across the Pharma value chain.

    Low dimensional representation learning of data that are naturally graph structured have gained popularity over recent years with varied applications. Graph neural networks, by design, are capable of integrating node informations, topological structure and relationship between data elements leading to increased accuracy of representing data from non-euclidian domains. Traditional deep learning representation typically struggles with tasks which require representing complex interdependence between data objects due to their inherent design of projecting embeddings in Euclidian space leading to the aggregation of relationships. GNNs however come in various flavors, graph convolution, temporal graphs, auto-encoders, transformers, spatio-temporal and others, which make GNNs quite versatile to tackle a diversity of use-cases frequently encountered in a large industrial setting. Here, we showcase our experiments using GNNs for low dimensional representation learning and using them to tackle various use cases across the pharma value chain. We compare and contrast these methods with more traditional ones and also highlight how such applications lead to generation of insights, and predictions with direct application towards various use cases.

    Srayanta is a Data Scientist and computational biologist with 10 years research experience, having worked a diverse spectrum of problems including predictive modeling and operations research.

    He has extensive experience in machine learning methods and is a specialist in stochastic simulations, deep learning and decision trees.

    His roles have included leading his team towards end-to-end data science solutions, achieved strategic milestones and drove adoption

    Linkedin
  • 09:40
    Varun Kohli

    Deep Learning to Detect Intrusion Attempts

    Varun Kohli - Lead Engineer - Machine Learning - Google

    Down arrow blue

    Deep Learning to detect Intrusion Attempts

    Machine learning and Artificial Intelligence are one of the hottest topics in computer science today. Many problems deemed ""impossible"" only five years ago have now been solved by machine learning: playing GO, recognizing what is in an image, or translating languages. Join us as we lift the lid on popular network attacks of 2022 and dive deep into the world of deep learning to understand how models can detect them proactively.

    Varun is a mountaineer and an engineer by heart, currently working as a machine learning engineer with Google and builds solutions to drive security insights for Alphabet entities. Along with his passion for high altitude mountain climbing, he is an avid machine learning developer, researcher and a public speaker. Google Crowdsource, Amdocs R&D, c0c0n, NASSCOM WWRT, Nullcon, and Nagarro DevOpsCon are to name a few.

    Twitter Linkedin
  • 10:05
    Sheraz Ahmed

    New Research into DL

    Sheraz Ahmed - Senior Researcher - DFKI

    Down arrow blue

    Ensuring Transparency

    Sheraz Ahmed is a Senior Researcher at Deutsches Forschungszentrum fur Kunstliche Intelligenz.

    He has worked for a variety of different research institutes including the University of Western Australia, Osaka Prefecture University and Fraunhofer ITWM

    Linkedin
  • 10:30

    COFFEE BREAK

  • DEEP LEARNING LANDSCAPES

  • 11:10
    Dubravko Dolic

    Visual Perception Provider: Is It Useful to Define you Own Deep Learning Framework?

    Dubravko Dolic - Head of Applied Analytics & AI - Continental

    Down arrow blue

    Visual Perception Provider: Is it useful to define your own Deep Learning Framework? A real life example from Continental Tires.

    At Continental Tires Deep Learning has its place in many areas. From Quality Assurance over Forecasting to Mobile Apps with Visual Elements. To enable Data Scientists creatin Deep Learning based applications quickly we defined a Framework to support those. How was it done and how does it help? What are the measures to ensure the business value created by such a framework. The Visual Perception Provider (VPP) is an inhouse solution build to solve problems in the Tire World.

    Programming with data since 1996. Focused on solving data driven problems. Generating insights by squeezing data sources. Loves IT and tools in the field of data analytics and data science.

    Linkedin
  • 11:35
    Samantha Edds

    Predicting Natural Disasters - Testing the Boundaries of Computer Vision

    Samantha Edds - Senior Data Scientist - Yelp

    Down arrow blue

    Predicting Natural Disasters- Testing the Boundaries of Computer Vision

    How well can a computer vision model trained only on point-of-view photos predict on aerial ones? Are some types and aspects of natural disasters easier to predict than others? Sam will speak about this, as well as the applicability and limitations of this kind of work to help us in the real world.

    Sam Edds is a passionate leader with a successful track record in using statistics and data modeling to help organizations uncover insights and tell a story to grow their business. Her unique background spanning corporation, start-up, and non-profit settings has shown me the importance of supporting the people, products, and places that make up a community. As a Statistician with roots in International Studies and Development, she firmly believes in harnessing the power of big data to improve the livelihood of all through making more informed, data-driven decisions. While there is more analysis than ever before in the world, something endlessly important to business success, and which remains her focus, is using big data to tell a story and a vision all can grasp. She loves designing and building models to solve problems, and thrives on using her analysis to create a story that all clients (data focused or otherwise) can understand.

    Twitter Linkedin
  • 12:00
    Johannes Hotter

    Open-Source Developer Toolkits for Building AI Training Data

    Johannes Hotter - Co-Founder / CEO - Kern.AI

    Down arrow blue

    Open-Source Developer Toolkits for Building AI Training Data

    Kern AI builds open-source developer toolkits for building AI training data. Their main product, refinery, is a development environment primarily for Natural Language, which enables its users to automatically build large-scale and high-quality training data both from scratch and from existing training data. With their tools, data-centric NLP models can be built within days instead of weeks.

    Linkedin
  • 12:25
    Hendrik Woerhle

    Deep Learning for Smart Living

    Hendrik Woerhle - Lecturer - University of Applied Sciences and Arts Dortmund

    Down arrow blue

    Deep Learning for Smart Living

    Hendrik is a Professor of Information Technology at the University of Applied Sciences and Art, and an expert in the application of artificial intelligence methods in embedded systems.

    Linkedin
  • 12:50

    LUNCH

  • MODEL ARCHITECTURE

  • 13:50
    Dr. Luis Moreira-Matias

    Eight Lessons from Creating ML-Based Products

    Dr. Luis Moreira-Matias - Director of Data & Machine Learning Engineer - Sennder

    Down arrow blue

    Eight Lessons From Creating ML-Based Products

    Nowadays, Machine Learning is an unrealistic hype in many industries. It is fun and relatively easy to design advanced machine learning (ML) as a fancy prototype in a simple laptop. However, the cold hard fact is that it is extremely hard to transform them into real but yet profitable software products. In this talk, I will disclose eight lessons that me and my teams learned on those journeys from notebook to production. Spoiler alert: yes, I will talk about data, maths, xAI, fairness, deep learning and all that jazz...but in the end, it is all about Engineering.

    Linkedin
  • 14:40
    Christoph Spohr

    DL at Volkswagen

    Christoph Spohr - Lead Architect - Volkswagen AG

    Down arrow blue

    Preparing your Data for DL

    Christoph Spohr is the Lead Architect of Big Data Platforms at Volkswagen following roles at both EPAM Systems and DATEV eG.

    Linkedin
  • 15:05

    COFFEE BREAK

  • DL CONSIDERATIONS

  • 15:50
    Yulia Grishina

    Deep Learning for Language Technology: The Science Behind Digital Assistants

    Yulia Grishina - Research Scientist - Amazon

    Down arrow blue

    Deep Learning for Language Technology: The Science Behind Digital Assistants

    This presentation focuses on the ML science and language technology that empowers Alexa AI. We will discuss emerging research questions that are faced by practitioners developing conversational assistants and in particular, we will dive deep into deep learning techniques that enable natural language understanding at scale.

    Yulia Grishina is a Research Scientist at Amazon Alexa AI in Berlin. She holds a PhD in the field of Natural Language Processing from the University of Potsdam, Germany.

    Linkedin
  • 16:15

    Panel: What are the Deep Learning Trends you Should Be Aware of?

  • Prokopis Gryllos

    PANELLIST

    Prokopis Gryllos - Senior Data Scientist - Shopify

    Down arrow blue

    Prokopis is a product-minded Data Scientist who enjoys Economics, Finance, and Algorithms

    Skills: data science, programming, product development, distributed systems Academic: statistics, machine learning, economics, game theory, social network analysis

    Linkedin
  • Rosona Eldred

    PANELLIST:

    Rosona Eldred - Machine Learning Engineer - BASF

    Down arrow blue

    Rosona is a Data Professional with 5 years of industry experience following an academic career in Mathematics culminating in a Max Planck research fellowship. Excels in collaborative teams with proactive independent contributors. Having worked with all parts of the ML life-cycle from requirements engineering to productionization, she is especially motivated by structural solutions to problems, by translating business potential to business value, getting promising prototypes effectively into production.

    Linkedin
  • Fabian Seipel

    PANELLIST

    Fabian Seipel - Lecturer - Deep Learning for Audio Event Detection - Technische Universitat Berlin

    Down arrow blue

    Fabian is interested in audio related research fields such as virtual acoustics, spatial audio, music information retrieval, digital signal processing and machine learning.

    Linkedin
  • 17:00

    NETWORKING RECEPTION

  • 18:00

    END OF DAY ONE

  • 08:00

    REGISTRATION & LIGHT BREAKFAST

  • 09:00

    WELCOME NOTE & OPENING REMARKS

  • APPLICATIONS IN DEEP LEARNING

  • 09:15
    Florin Coman

    Top 3 Most Complex Projects in the Current Environment

    Florin Coman - Conversational AI Architect   - Bosch

    Down arrow blue

    Top 3 Most Complex Projects in the Current Environment

    Even if AI provides “state of the art” solutions and research is gaining more ground, in real life Conversational AI doesn’t always have the impact expected by many. Florin and Zsolt Szekely will address the current challenges and will present a live demo of the latest and most complex developments. We at Bosch Service Solutions believe that the time for power point presentations, in which we create unrealistic expectations, need to be put behind us, and the focus should be on projects that can have immediate impact and the capacity to engage users, giving the same importance also to low-resourced languages.

    This year, Florin together with his team started the implementation of complex Conversational AI frameworks (available for all languages, channels, and use-cases) designed to improve the AI based solutions within Bosch SO. The target is to continue the deployment of projects in each of the five Conversational AI pillars: Voicebots, Chatbots, Translation AI, Text/Email AI and Agent Assist AI. Prior to this role, Florin was involved in research, start-ups and other summits as keynote speaker - all these experiences giving him a 360 degree perspective on trends and challenges.

    Linkedin
  • 09:40
    Han Xiao

    Unlock the Business Value of your Multimodal Data via Jina AI

    Han Xiao - Founder & CEO - Jina AI

    Down arrow blue

    Unlock the Business Value of your Multimodal Data via Jina AI

    Multimodal AI is a new type of AI that can take in data from multiple modalities (images, text, audio, etc.) and learn from it to make predictions. This makes it very powerful for applications like neural search and creative AI (DALLE and StableDiffusion). Multimodal AI also presents new challenges for engineering. In particular, it can be difficult to deploy and manage multimodal AI models and pipelines and make them scalable in production. This is where MLOps platforms come in. In this talk, we will discuss the benefits of using an MLOps platform for multimodal AI applications in production and help to unlock the business value of your multimodal data.

    Dr. Han Xiao is the Founder & CEO of Jina AI, a commercial opensource company based in Berlin. Since its founding in 2020, Jina AI has raised $38M from top investors, including GGV, Cannan, YUNQI, SAP.io. Jina AI is one of the most promising AI startups globally according to CBInsights 2022, 2021 and Forbes DACH 2020. Before Jina AI, Han led a team on neural information retrieval at Tencent AI, laying down the next-gen search infrastructure. Han served as a board member at Linux Foundation AI in 2019, driving the opensource innovation and international collaboration. In 2014-18 Han worked at Zalando Research in Berlin as a Senior Research Scientist. Han received a Ph.D. (2014) and MSc. (2011) in computer science from the Technical University of Munich in Germany.

    Linkedin
  • 09:55
    Aleksandra Kovachev

    Dish Catalogue Optimisation: From Unstructured Data to Knowledge Graph

    Aleksandra Kovachev - Data Science Manager - Delivery Hero

    Down arrow blue

    Dish Catalogue Optimisation: From Unstructured data to knowledge graph

    Can you imagine the beauty of the data we work with at Delivery Hero? We hold knowledge of the world’s food culture in 74 countries, as of 2022 including Glovo! To extract the hidden knowledge that our data holds, we need to transform the free and unstructured data into a meaningful catalogue and taxonomy. In this talk we will show you how we use state-of-the-art machine learning models to provide structure, relationship and meaning to our products worldwide. We will also focus on the main challenges, learnings and solutions we applied tackling this complex task, not only from a data science point, but also from a product perspective. Making this data available to our stakeholders at hand we are able to facilitate dominance in innovation, quality and customer satisfaction in our business

    Aleksandra did her PhD in the area of complex networks with the goal of knowledge extraction by combining multiple data sources and diverse algorithms. She has passion in bioinformatics and improving health trough food and nutrition data. Currently she works as ML Engineer for the global food delivery service, Delivery Hero.

    Linkedin
  • 10:20
    Özlem Gürses

    The Concept of Fraud in Machine Learning

    Özlem Gürses - Professor - Kings College London

    Down arrow blue

    The legal Concept of Fraud in Machine Learning

    In legal terms fraud is not mere lying; it is seeking to obtain advantage, usually monetary, or to put someone else at a disadvantage by lies and deceit. In order to prove a person has acted fraudulently, it is necessary prove that that person was either deliberately or recklessly intended to defraud. There are fine lines to be drawn between 'negligent', 'reckless' and 'deliberate' acts and only the last two suffice to prove fraud. Whilst training machines to detect 'fraud' such fine lines must be observed. Otherwise, framing an action as a 'fraud' whereas there was no intention to 'defraud' might have serious consequences especially for consumers and the relevant actor who is responsible for such misclassification might be found liable to compensate losses suffered by many consumers who have been affected by it. I therefore look forward to discovering more at Berlin Summit how experts define and design 'fraud detection' in their machine training processes.

    Özlem Gürses is Professor of Commercial Law at King’s College London. She specialises in insurance and reinsurance law. Özlem is the author of Reinsuring Clauses (Informa), Marine Insurance Law (Routledge), Insurance of Commercial Risks (Sweet and Maxwell), and The Compulsory Motor Vehicle Insurance (Informa) as well as numerous articles published on insurance and reinsurance related topics. Özlem sits in the British Insurance Law Association Committee and the Presidential Council of the International Insurance Law Association (AIDA). She is Vice-Chair of the Reinsurance Working Party of AIDA. Özlem teaches insurance and reinsurance law at King’s College London and abroad, including National University of Singapore, University of Hamburg and World Maritime University, Malmö

  • 10:45

    COFFEE BREAK

  • NATURAL LANGUAGE PROCESSING

  • 11:25
    Roshan Amasa

    We Give it to the Projects, Spread it Across the Countries!" - NLP at Munich Re.

    Roshan Amasa - Lead Solutions Architect - Data & AI - Munich Re

    Down arrow blue

    We Give it to the Projects, Spread it Across the Countries!" - NLP at Munich Re.

    • Since 2018, Munich Re is developing an in-house NLP platform to quickly transfer research insights into production-ready models. While it is quite common that knowledge gets lost from one project to another, we systematically capture reusable components in our own library, standardize and share code and models across projects for reuse and collaboration. This facilitates the creation of new solutions for any of our companies.

    • Roshan works as a Lead Architect for Business Technology at Munich Re and consults various projects including the NLP platform. Prior, he worked in the automobile industry handling Big Data, MLOps and data platform topics. He holds a Master’s degree in Software Engineering. He is passionate about serverless technologies and loves helping organizations in their data, AI and cloud journeys .

    Linkedin
  • 11:50
    Virendra Kumar Pathak

    Applying NLP in the Automotive Domain and Challenges Involved.

    Virendra Kumar Pathak - NLP & Deep Learning Engineer - BMW

    Down arrow blue

    Applying NLP in the Automotive Domain and Challenges Involved

    Virendra is an NLP and Deep Learning engineer at the Artificial Intelligence team of BMW Group. At BMW, he primarily focuses on industrializing AI in the quality domain. He takes particular interest in transforming business requirements into AI problems and using semi-supervised approaches, especially in the absence of labeled data.

    He moved to Germany for his Master's in Informatics at TU Munich. Previously, he worked in the semiconductor industry, building Linux drivers and compiler optimizations for HPC processors.

    Linkedin
  • TOOLS FOR DEEP LEARNING

  • 12:15
    Nima Siboni

    Reinforcement learning in Transport and Logistics

    Nima Siboni - Researcher - Max-Planck-Institute

    Down arrow blue

    Reinforcement Learning in Transport and Logistics

    This presentation is dedicated to the challenges and successes of utilizing reinforcement learning as an optimization tool in transport.

    Nima is a Researcher for Machine Learning at Max Planck Institute and Senior RL Research Engineer at InstaDeep Ltd. He is an AI-practitioner and experienced Simulation Scientist with focus on Complex Systems.

    Linkedin
  • 12:40

    LUNCH

  • 13:40

    RESERVED FOR OVERLEAF

  • 13:55
    Sergei Bobrovskyi

    Real Time Supervised Anomaly Detection

    Sergei Bobrovskyi - Expert Anomaly Detection - Airbus

    Down arrow blue

    Real Time Supervised Anomaly Detection

    Anomalies are patterns in data not conforming to expected behavior. Discovery of such patterns leads to actionable insights as anomalies often correspond to undesired states e.g. reduced quality or some failure. While automatic finding of anomalies is a challenge in itself, it is usually followed up by another hard task to understand the causes of their occurrence, in order to prevent them from happening again. The possibility to explain the decision of the anomaly detection system not only helps to establish trust but also to identify the root causes. In this talk I investigate this connection and challenges accompanying Deep Learning approaches.

    Dr. Sergei Bobrovskyi is a Data Scientist within the Analytics Accelerator team of the Airbus Digital Transformation Office. His work focuses on applications of AI for anomaly detection in time series, spanning various use-cases across Airbus. Prior to Airbus he worked on automated fraud detection for one of the largest e-commerce companies in Germany. Before that he was engaged in various research related positions in the space industry.

    Sergei holds a PhD in theoretical physics as well as a physics Diploma from the University of Hamburg. Besides physics he also studied philosophy with an emphasis on the philosophy of mind.

    Twitter Linkedin
  • 14:20

    Panel: The ROI of DL

  • 15:00

    END OF SUMMIT

Berlin AI Summit

Berlin AI Summit

04 - 05 October 2022

Get your ticket
This website uses cookies to ensure you get the best experience. Learn more