REGISTRATION | Registration, coffee, networking: 8:30-9:30 |
OPENING NOTES - (Norbert Végh - EPAM) | Opening notes by Norbert Vegh, Country Head of EPAM Hungary. |
KEYNOTE: Google - Using GenAI with Python to understand and edit multimedia - (Markku Lepistö, Google) | In this session Markku from Google will show how to use GenAI services from Python, to work with media content. The goals include to
- understand the context of images
- selectively and generatively edit objects in images
- chain multiple GenAI services to e.g help visually impaired people understand their surroundings, by hearing a verbal description of the scene in front of them
The session includes live demos and explanation of the main code parts |
KEYNOTE: Microsoft - Navigating the AI Revolution: Microsoft’s Vision and Tools for Developers - (József Zoltán Vadkerti, PMP, Microsoft) | In this exciting keynote, we’ll journey through the AI era, highlighting Microsoft’s innovative AI strategy. We’ll explore the latest AI tools for developers, such as Azure AI Studio and Azure Machine Learning Prompt Flow. We’ll introduce GitHub Copilot, your AI pair programmer that’s reshaping the coding landscape. Plus, we’ll delve into the rise of low-code/no-code development with Microsoft’s Power Platform. This presentation is a must-attend for developers eager to stay at the forefront of the AI revolution. |
KEYNOTE: Amazon Web Services - AWS: closer to the future - (Krisztián Szulák - AWS) | In this talk you can learn about current trends in the Cloud and what AWS can offer you to keep up with the ever-changing world of IT. AWS is constantly extending it's presence in the European Union and also in Hungary. Let's see how close are we to the Cloud. |
BREAK | Break - 30 minutes |
ROOM 1: Unleash Your Coding Potential with GitHub Copilot - (József Zoltán Vadkerti, PMP, Microsoft) | Join us for an exciting demo where we’ll harness the power of GitHub Copilot in a .NET Core environment! We’ll build a console application, explore real-time code completions with GitHub Copilot Chat, and dive into the GitHub Copilot Lab in Visual Studio Code and Visual Studio.But that’s not all! We’ll also tackle critical aspects of software development like writing unit tests, squashing bugs, and boosting performance, all with the help of GitHub Copilot. This demo is set to revolutionize your coding experience, increase productivity, and reduce development time. Don’t miss out! |
ROOM 2: Gen AI as an opportunity for your business - (Pavel Krejsa - Google) | How Gen AI changes your business by improving the life of your Data analysts. Let's discover applications of Gen AI in Google BigQuery analytical database and how this does improve your Data analyst daily job and reduce toil. We will see practical examples from the world of SQL. |
BREAK | Break - 10 minutes |
ROOM 1: Amazon Bedrock - (Árpád Csőke - AWS) | The hottest topic this year is undeniably Generative AI. In this talk you will learn about the tools and services AWS is bringing to the table and how you can use them in different scenarios. |
ROOM 2: Vertex Search & Conversation - (Antal Mihályi, Google) | We will talk about Foundation models and how to build apps with them. Let's discuss the foundation models you can build on, what are the challenges of LLM based application development, and what tools Vertex can provide to overcome. |
BREAK | Break - 10 minutes |
ROOM 1: Intro to Model Garden and GenApps on Google Cloud - (Márton Kodok - Google Developer Expert, Reea) | Build applications with generative AI on Google Cloud. Vertex AI is a managed unified ML platform for all your AI workloads on Google Cloud Platform. We are going to cover what Gen App Builder is for developers to build and deploy AI-driven applications. We will explore Model Garden powered experiences from PaLM to foundation models, then we are going to learn more about the integration of these generative AI APIs. At the end of the session, developers will understand how to innovate with generative AI and develop apps using the generative ai industry trends. |
ROOM 2: Efficiency in Motion: Practical Tips for Autoscaling in Kubernetes - (Yurii Kyrychynskyi - Kiwi.com) | In this talk, I would like to share tips and tools to boost Kubernetes autoscaling. At Kiwi.com, we extensively leverage the Google Cloud Platform (GCP), and most of the infrastructure runs in GKE. I will describe the scaling problems we hit and how tools such as KEDA and Descheduler helped us to overcome them. These tools played pivotal roles in improving elasticity, minimizing costs, and enhancing the overall performance of our Kubernetes deployments. |
LUNCH | Lunch, networking |
ROOM 1: Streamlining AI Application Front-Ends: It’s time to think about fast AI apps! - Susanna Wong - (Google Developer Expert, QuantumBlack) | In today's AI-driven world, integrating AI technologies into web apps is the future. However, it is a challenge to ensure a seamless and responsive user experience while dealing with AI processing. As a FE engineer building web tools for the ML space, I will explore the strategies and best practices for architecting your FE app to allow fast and seamless AI integration, diving into the set of core techniques for optimizing loading times, such as minimizing HTTP requests, asynchronous loading, leveraging caching, and optimizing assets like images. We will also discuss the benefits of server-side rendering (SSR), code splitting, and the use of web workers to offload AI tasks. The talk will emphasize the significance of delivering AI-driven features without compromising speed, accessibility, or user satisfaction. By the end, the audience should have a comprehensive understanding of how to architect and optimize AI applications effectively in striking a harmonious balance between cutting-edge AI functionalities and an exceptional user experience. Let’s empower the future set of AI applications with the power of the fast web! |
ROOM 2: Crossplane and a story about scaling Kubernetes custom resources - (Richárd Kovács - Akamai) | In the ever-evolving landscape of cloud-native technologies, Kubernetes has emerged as the de facto container orchestration platform. With its vast ecosystem of extensions, the ability to customize Kubernetes resources is essential for organizations seeking to harness its full potential. This presentation explores the fascinating world of Crossplane, an open-source project that empowers users to scale Kubernetes custom resources to new heights. Crossplane is a revolutionary tool that bridges the gap between Kubernetes and cloud infrastructure, enabling users to define and manage custom resources that seamlessly integrate with their cloud services and applications. But what makes this journey even more intriguing is the narrative of scaling these custom resources. In this presentation, we'll embark on a storytelling journey that illustrates real-world scenarios and experiences of organizations as they leverage Crossplane to scale their Kubernetes custom resources. From the inception of custom resource definitions (CRDs) to managing complex, distributed applications at scale, we'll delve into the challenges faced, innovative solutions deployed, and the successes achieved. Join us as we unravel the tale of organizations who've used Crossplane to create, manage, and scale custom resources in Kubernetes, unlocking a new level of automation, flexibility, and efficiency. Discover how Crossplane is transforming the Kubernetes landscape and learn from these inspiring stories about scaling custom resources to drive innovation and growth in the cloud-native era. |
BREAK | Break - 10 minutes |
ROOM 1: Discovering PaLM - Aqsa Kausar - (Google Developer Expert, Red Buffer) | The talk will be covering GenAI studio in Google Cloud and particulary the language component where we will see how PaLM can be used to generate text. We will also look into how we can use the different hyperparameters and the effect they have. |
ROOM 2: Efficient Observability With Telemetry Pipelines - (Sándor Guba - Axoflow) | Metrics, logs, and traces are the bread and butter for operations and developers. The problem is that observability data continues to grow yearly. Transporting, storing, and extracting value from it can be expensive. This is the reason behind the latest trend “Telemetry Pipelines” (also called “Observability Pipelines”). They help to collect, enrich, filter, and transport observability data. The more you prepare your data the better results you will get in your SIEM or observability systems. Sandor will demonstrate the strength of such a system throught the Kubernetes Logging Operator /CNCF Sandbox project/. |
BREAK | Break - 30 minutes |
ROOM 1: EPAM Keynote - AI DIAL - Unified AI Access for Enterprises - (Ilya Gorelik - EPAM) | AI DIAL is a secure, scalable and customizable enterprise-grade AI ecosystem that seamlessly integrates with your data and workflows, tailored to achieve your unique business objectives. This talk is a hands-on demonstration of AI DIAL by Ilya Gorelik, Head of Reliable AI (RAIL) and Real Time Computing (RTC) Labs, EPAM Systems. |
ROOM 2: Challenges of ML-based IoT Malware Detection - (Tatiana Barbova - Cloudera) | Let's have an overview of how ML came into the domain of malware detection, what are the current approaches and their performance (from Random Forests to CNNs), with a particular focus in IoT malware detection. Every stage of the ML pipeline for this problem will involve challenges and decision tradeoffs, so I would like to present you practical solutions based on the latest research in the field. |
Break | Break - 10 minutes |
ROOM 1: Practical AI Masterclass in the biggest European OpenStack Cloud - (Ferenc Kukucska - Deutsche Telekom IT Solutions) | In this session Kukucska Ferenc, the member of the Big Data & AI Squad of Open Telekom Cloud will give insights, how they have built a Plant Disease Detector web app and mobile app, which can help define the type of a plant based on one single leaf. This application also gives the opportunity for users to diagnose, if the plant is healthy or unhealthy including the disease type as well.
The focus will be on the AI part of the app but the following will be also covered: How the infrastructure of the Plant Disease application was built, describing the whole pipeline from uploading the dataset to the OTC Object Storage Service (OBS) through training the model in Open Telekom Cloud's AI service Modelarts (MA) till the Kubernetes deployment and secret management of the whole application describing each technical step in great details. |
ROOM 2: Backups in the Kubernetes era - (Zsolt Keserű - EPAM) | Backup/restore process expectations vs. reality in the world of cloud native applications. |
BREAK | Break - 30 minutes |
ROOM 1: AI in Cybersecurity - (Ádám Balázs - EPAM) | AI have entered the Cybersecurity game in 2020 and its revenue was over $10 billion, now experts say in the next few years it can be even 5x higher than that. But what is behind those numbers? How do big tech companies such as Microsoft, SentinelOne, Google or even Mastercard use AI in Cybersecurity? It is to be found out. |
ROOM 2: Sustainability for workloads in Google Cloud - (Irine Kokilashvili - EPAM) | Sustainability in the cloud is the continuous dedication to primarily decrease energy usage and optimize efficiency across all aspects of workloads. Assuming the responsibility for your emissions embodies ethical values, cost-efficiency, regulatory compliance, and environmental impact awareness.
It also reflects your commitment to being the part of the solution rather than contributing to the problem. In this talk you will learn about design principles, best-practices, potential trade-offs, and improvement plans you can use to meet sustainability targets for your Google Cloud workloads. |
BREAK | Break - 10 minutes |
ROOM 1: Accelerate front-end development with Bard and IDX - (Teresa Wu - Google Developer Expert, J.P. Morgan) | In this talk, I will give a brief overview of a few keywords in the AI and Machine learning world, such as types of ML, LLM, NLP, generative, GPT, and prompts, which then leads into the conversation of how AI helps engineers to work better, giving a demo of making a front-end application using Bard with is powered by PaML2.Key takeaways:
- Demo the process of creating a front-end project using Bard
- Understand how AI can accelerate the project development and release process
- No-code and Low-code platforms
- The benefit of using IDX - a AI-powered IDE to accelerate the development process |
ROOM 2: Rapid product prototyping with managed GCP services (Firebase) - (Olivér Lutz, Márk Rákóczi, Krisztián Dobos - Deutsche Telekom IT Solutions) | From an idea to a fully working prototype within weeks? Is it possible? We show you how we created a small product for internal use within a couple of weeks with a small team of junior dev, whom are mainly infra engineers (has no deep coding experience) and how easily can you prototype your idea as well with cloud managed services (Firebase) and openly available building blocks. |
BURGER & BEER / NETWORKING | Beer, burger, networking |