AI Meetups 🇺🇦

Hi, guys! Let us tell you about our first meetup in Kyiv this year 🎯 We were visited by Andrii Polukhin, ML Tech Lead at Samba TV, with the topic “Transforming ML Development with Multimodal Zero-Shot Models” We talked a lot about embeddings and their combination in one space, held a cool Q&A session, and talked to the speaker behind the scenes 😎 This time, we managed to gather about 60 participants offline and online! We are grateful to Andrii for his extraordinary efforts and to Kyiv School of Economics for providing the space for our meeting and helping to organize the event! For those who were unable to join, the recording of the meeting and materials are available at the link - https://lnkd.in/dJUvTj_r We look forward to seeing you at our next events! Follow the announcements in our channel and on our Telegram (https://lnkd.in/dBEHW6M2) and Instagram (https://lnkd.in/d4sQhfT4) pages!

Last weekend, I had the opportunity to speak at the Kyiv School of Economics on the topic of Transforming ML Development with Multimodal Zero-Shot Models. Big thanks to AI Meetups 🇺🇦 for the invitation and great organization! We explored how multimodal models unify different data domains through shared embeddings, dived into feature alignment and token-sequence modeling, and discussed how vision-language models (VLMs) are built. Thank you to everyone who attended in person and joined us online! I will share the recording and lecture materials later — stay tuned! 🚀 #ai #vlm #ml #kse #multimodal

AI Meetups - Transforming ML Development with Multimodal Zero-Shot Models We invite you to join our meetup in Kyiv! 🔥 🧑💻Speaker: Andrii Polukhin - ML Tech Lead @ Samba TV, PhD student, Lecturer & Mentor At the meetup, we will discuss how multimodal models unify diverse data domains using shared embeddings. We will dive into the mechanisms of feature alignment and token-sequence modeling, and consider how to improve dataset annotation with VLM. Topics to be covered: 1. Explore how multimodal models unify diverse data domains (image, text, video). 2. Practical delve into feature alignment, token-sequence modeling. 3. Code-driven methods for dataset enhancement and cross-modal task. 4. Refining annotations and generate additional data with vision-language models. When: Saturday, January 25 Schedule:  18:00 - 19:00: Lecture  19:00 - 19:30: Q&A  19:30+: Networking 📌 Where? 📍 Kyiv, Kyiv School of Economics 📍 Online Registration: https://lnkd.in/dyVhf9aW

2024 is coming to an end, so let’s take a moment to reflect on what an incredible year it has been 🙌🏻 Looking back, we’re amazed that we managed to organize 15 meetups and bring together over 700 people in just one year! Our events took place in Lviv and Kyiv, and online for anyone who wants to join👨🏽💻 Are we planning to host even more exciting meetups and engage even more people next year? Absolutely 🤍 Thank you for being with us ❤️ See you in the upcoming 2025!

AI Meetups - The Future of Human-Computer Interaction We invite you to join our upcoming meetup in Lviv! Explore a future where tasks can be delegated to autonomous agents, the models currently used to build them, and the accessibility and understanding of user interfaces. Speaker: Mariya Hirna, Data Scientist @ MacPaw 🇺🇦, UCL & UCU Faculty of Applied Sciences / APPS UCU Alumni Topics to be covered: 1. Race for automatic computer agents: are we there yet? 2. UI representation: user interface as a data modality 3. Automatic data collection for computer task generation 4. VLMs and agentic systems for automatic task execution Date and time: Saturday, December 7 ⏰ 18:00 - 19:00: Lecture ⏰ 19:00 - 19:30: Q&A Session ⏰ 19:30+: Networking Venue: 📍Online 📍Lviv, Ukrainian Catholic University (details upon registration) Registration: https://lnkd.in/dwW--_BE See you there! 🚀

Hello, everyone! Sharing our impressions from the last AI Meetups in Kyiv! 🎯 This time, we had the honor to host Volodymyr Sydorskyi on the topic of “Skin Cancer Detection Kaggle Competition ISIC 2024” We managed to gather about 60 offline and online participants for the event this time! We sincerely thank Volodymyr for an extremely informative lecture and Kyiv School of Economics for providing the space for our meeting and helping to organize it! For those who were unable to join us, the recording of the lecture is available here - https://lnkd.in/du8Wwbem We look forward to seeing you at our next events! Keep up to date with event announcements at our Telegram (https://lnkd.in/dBEHW6M2) and Instagram (https://lnkd.in/d4sQhfT4) pages

We are pleased to invite you to join our meetup in Kyiv! 🔥 🧑💻 Speaker: Volodymyr Sydorskyi - Computer Vision Engineer at Better Medicine, Kaggle Competitions GrandMaster, Co-Founder of Zvook, PhD Student at IASA, finalist of Kaggle Competition ISIC 2024 At the event, we will discuss tips & tricks for achieving high results at Kaggle competitions. Also we will find out whether Kaggle models can be applied to real patients and why decision trees can outperform CNNs in image processing tasks. Topics to be covered: 1. Falls and Rises on Kaggle or How to keep mental health in Competitive ML 2. Can we use you Kaggle models on real patients? 3. Decision Trees outperform CNNs on Imaging task. WAT? 4. Validation is all you need 5. Stack more models! 📆 When? Saturday, November 9 ⏰ Schedule:  18:00 - 19:00: Lecture  19:00 - 19:30: Q&A  19:30+: Networking 📌 Where? 📍 Kyiv, Kyiv School of Economics  📍 Online Registration - https://lnkd.in/dKVpwqAN

Recapping our recent meetup featuring Bogdan Didenko! 🎉 This time, we brought together over 50 attendees, both in Lviv and online! 🥳 For those who couldn't make it, here's the recording of our meetup: https://lnkd.in/gFr6rT_c A big shoutout to Bogdan for the great talk, and to the hosting venue, UCU Faculty of Applied Sciences / APPS UCU, for providing us with the space to gather and learn. We look forward to seeing you at our future meetups! 🤝 Stay tuned for updates on our upcoming meetups by following us on this page or joining our Telegram: https://lnkd.in/dBEHW6M2

AI Meetups - Solving Overlooked NLP Problem with RedPenNet Architecture We invite you to join our September meetup in Lviv with Bogdan Didenko, Deep Learning Engineer @ WebSpellChecker At the meetup, we'll discuss the specifics of the Sequence-To-Edits task, its difference from Sequence-To-Sequence and Sequence Tagging. Bogdan will talk about the RedPenNet architecture he developed to solve the sequence editing task, which uses the attention mechanism to find positions in the text Topics to be Covered: 1. Why ChatGPT regenerate entire previous responses even when asked for minor edits 2. Limitations of autoregressive generation and sequence tagging in solving certain tasks 3. Semi-autoregressive approaches 4. Utilization of the attention mechanism in the Pointer Networks approach 📆 When? Wednesday, September 25 ⏰ Agenda: 19:00 - 20:00: Lecture 20:00 - 20:30: Q&A Session 20:30+: Networking 📌 Where? 📍 Lviv, Ukrainian Catholic University 📍 Online 🎟 Secure your spot now: https://lnkd.in/dwBcz6vw

AI Meetups - Architectural Automation with AI Excited to see you at our Kyiv meetup! 🔥 Featured Speaker:  David Agafonov – Solution Architect in VR/AR, AI, CTO @ ByteAI.io What’s on the Agenda:  Discover the latest in architectural automation using IQaaS, no-code, and low-code platforms. Learn about chain modules, Langchain, and Vellum.ai for efficient prototyping and LLM interaction. We’ll also delve into prompt engineering challenges and their effects on code quality and development speed. Key Topics: 1. Building a Low-Code Architecture Planning App 2. ByteAI’s Initial Prototype: Chain Modules and Files 3. Enhancing & Automating Prototyping: Langchain & Vellum.ai 4. Data Preparation: Ensuring Effective LLM Interaction 5. Prompt Engineering: Common Pitfalls and Best Practices Date: Saturday, September 28 Schedule: 18:00 - 19:00: Lecture 19:00 - 19:30: Q&A 19:30+: Networking Location: 📍 Kyiv, Kyiv School of Economics 📍 Online Register Here: https://lnkd.in/dJ_qb_pE