Building Useful AI: From LLM Agents to Financial Products

AI is everywhere, but what does it actually look like when it works? Join us for a meetup that brings together practitioners building and deploying AI systems in real-world environments.

⚡ Building Smarter Financial Technologies with AI

This session will cover practical AI solutions developed for financial institutions, including:

• AI-powered data warehouse and AI Data Analyst
• AI chatbot for customer support
• Segmentation tools and audio analysis for call centers
• Income analysis tools for self-employed customers

Speakers:

Aleksi Kvitatiani - AI & Data Project Lead at Silk Bank, where he built and manages a cross-functional team responsible for the bank's end-to-end data ecosystem. His work includes an AI-powered data warehouse that lets business users query data in natural language, an agentic chatbot that reduced call center load by 75%, and an AI-assisted income analyzer for credit risk teams. Before Silk Bank, he consulted as a freelance data analyst for FMCG operations in Georgia. He also co-founded Outswift, a studio building ML and agentic AI solutions for SMBs.

Shalva Elishakovi - AI, Data & Growth Project Manager at Silk Bank, where he leads the development of AI-driven solutions across business functions. He has contributed to building one of the first AI-powered banking assistants with secure access to customer data and analytical capabilities for improved support and decision-making.
He is also a Co-Founder of Outswift, developing AI products across industries, including computer vision for FMCG and audio intelligence for call center analysis, with a focus on scalable, real-world solutions.

⚡ It turns out you do not need a multi-agent system

Ever thought about building a multi-agent system just because business or marketing said it was a cool idea? Well, we all have.

Over the past two years, we’ve been sold countless ideas and swept up in plenty of hype. But as things are becoming somewhat “production-ready”— thanks to coding assistants emerging as the most practical use case for LLMs — we’re now seeing Agent Skills and sub-agents as more effective patterns for building complex systems.

Let’s dive into the recent trends that are sticking and explore the challenges we still face, including token consumption costs, guardrails, and creating agents that truly deliver customer value.

Speaker:

Natig Kurbanov - Lead Software Engineer and an AI Solutions Engineer at EPAM Georgia. As an EPAM AI Ambassador and community driver for JAR.GE and GenAI-X Georgia, he is a frequent face on the tech circuit. During the networking part, feel free to go beyond Natig’s talk theme of multi-agent system — ask him about Japanese maples, magnolias, or Zen gardens, and you may find yourself in a long and engaging conversation.

When: 7 May, 18:30
Where: EPAM Georgia (10 Bidzina Kvernadze Str., Tbilisi)
Attendance is free and requires RSVP on Meetup.com.

🤝 The Meetup is organised by DataFest Tbilisi and hosted by EPAM Georgia. The event partners are World Quant and Charte.