SoftServe Poland

The event is free, register now 👉 https://hubs.ly/Q02Z8_FT0

📊 What’s on the agenda?
17:30 – 17:50 Come and grab a drink!
18:00 – 18:10 Official Start
18:10 – 18:30 Let’s break the ice 🤝
18:30 – 19:00 Talk#1 “Charting and using Knowledge Graphs with GPT: a Complete Journey” by Daniel Einor, Senior Research Analyst
19:00 – 19:30 Networking and Pizza
19:30 – 20:00 Talk#2 “Databricks—not only Spark on steroids” by Piotr Uzar, Data & Analytics Solutions Architect

✔️ Talk#1 “Charting and using Knowledge Graphs with GPT: a Complete Journey” by Daniel Einor, Senior Research Analyst
This talk explores the building and utilizing GPT-powered Knowledge Graphs, from crafting custom ontologies to implementing automated graph creation and beyond. Learn about common pitfalls in GPT-based KG agents and see practical approaches for managing graph databases at scale. Discover how combining GPT's capabilities with Knowledge Graph structures lets to create robust information systems that maintain data integrity while leveraging the power of Large Language Models.

✔️ Talk#2 “Databricks—not only Spark on steroids” by Piotr Uzar, Data & Analytics Solutions Architect
For many years, Databricks was considered as Managed Spark offered by cloud providers, and that was true. Join the presentation and discover the features like Unity Catalog, SQL warehouse, LakeFlow, DBT, serverless approach, and more that now allow us to think about Databricks as (almost) a complete solution for addressing a number of requirements, and allowing us to span multiple clouds.

⚡ Save your seat now! https://hubs.ly/Q02Z8_FT0 ⚡

🗣️ About speakers:
Daniel Einor, Senior Research Analyst, an avid researcher with a passion for observing patterns in nature and gaming systems. In the past year, blended traditional machine learning expertise with Knowledge Graph understanding and LLM architectures, drawing insights from his technical-naturalistic background. Currently focused on bridging the gap between human knowledge and machine learning capabilities.

Piotr Uzar, Data & Analytics Solutions Architect, a Solution Architect with 20+ years of experience in IT. He has held various roles: Network Administrator, Network Architect, Linux admin, Cloud Administrator and Cloud Architect, Enterprise Architect - Big Data Engineer and Data Architect. A new technology enthusiast and automation fan, Databricks and Azure promoter. Certification enjoyer. He channels his skills into real-world automation projects, such as using a Raspberry Pi to automate garden watering.

Save your seat now 👉 https://hubs.ly/Q02zDdB20

📊 What’s on the agenda?
17:30 – 17:50 Come and grab a drink!
18:00 – 18:10 Official Start
18:10 – 18:30 Let’s break the ice 🤝
18:30 – 19:00 Talk#1 "The DOs and DONt's of LLM Supervised Fine-Tuning" by Vladyslava Tyshchenko, Data Scientist
19:00 – 19:30 Networking and Pizza
19:30 – 20:00 Talk#2 "Apache Iceberg in modern Data Lake" by Michał Mroczek, Senior Big Data Software Engineer

Talk#1 “The DOs and DONt's of LLM Supervised Fine-Tuning" by Vladyslava Tyshchenko
In this talk, we will explore various peculiarities of supervised fine-tuning of LLM and the toolbox needed to accomplish it based on real examples successfully. We will cover the main components required to teach the LLM to follow instructions better, including the dataset, model selection, adapters, quantization, and more. If you are keen to learn how to create your own specialized LLM - join this session!
Talk#2 "Apache Iceberg in modern Data Lake" by Michał Mroczek
During the session, we'll explore Apache Iceberg, a high-performance table format for massive analytic datasets that enables ACID transactions, time travel, and seamless schema evolution in modern data lakes.

⚡ It is free of charge, but registration is required ⚡

Discover more about our speakers:

Vladyslava Tyshchenko - Vladyslava is a Senior Data Scientist at SoftServe. She has experience in building Machine Learning solutions and leading teams starting from business ideas to successfully deployed applications. Her knowledge spans across the field of Natural Language Processing back to the days when Transformer architecture was not yet invented. In her daily work she mostly builds LLM-based solutions for educational companies and institutions providing students with personalized educational experiences.

Michał Mroczek - Data and Cloud Engineer who transforms business challenges into scalable AWS solutions, specializing in data architecture and cloud infrastructure. With several years of hands-on experience, they excel in designing ETL pipelines, implementing and managing infrastructure as code. His expertise spans both technical implementation and team collaboration, consistently delivering data-driven solutions while maintaining robust security and governance standards.