#GenAIBootcamp by Disrupt AI Summit

April 3, 2026 - OTEAcademy Athens, Greece

A special one-day in person training, designed for Tech leaders
interested to upgrade their AI skills

How it Works

This bootcamp runs alongside, Disrupt AI Summit, focusing on the skills tech leaders and engineers need to acquire in order to thrive in the AI era.

In six hours attendees will get hands on actionable insights on artificial intelligence followed interactive workshop sessions. The training event capacity will be limited – up to 100 attendees

#GenAIBootcamp Agenda

"From Zero to Hero in One Day"

An immersive, hands-on technical bootcamp designed for engineers who want to dive into the AI era. Participants will learn LLM fundamentals, set up local models (Ollama, LM Studio), and build their first generative AI web application using TypeScript, Vercel AI SDK, Next.js, and Shadcn UI components. The session covers Retrieval-Augmented Generation (RAG) and tool integration, empowering attendees to code and deploy their own GenAI apps in one day—all running on their local machines.

By day’s end, every participant will have a working AI application running on their local machine—ready to extend and deploy to production.

Intro on AI Product Management

Introduction to Product Management, AI Concepts and how these two are connected to create this hybrid role.

09:00–09:30 | Unified Kickoff (30')

Welcome & bootcamp objectives alignment
AI era context: LLMs transforming industries
Bootcamp goals: Build, deploy, and showcase a GenAI application
Participant expectations & learning outcomes
Logistics: Discord channels, VS Code setup, resource links

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

09:30–10:30 | LLM Fundamentals Deep Dive (60 min)

What are Large Language Models? Architecture overview
Transformer architecture basics (attention mechanisms)
How LLMs generate text: tokenization, inference, sampling
Key concepts: tokens, context windows, temperature, top-p sampling
Prompt engineering principles: clarity, specificity, examples
Use cases: summarization, Q&A, code generation, creative writing
Model selection criteria: size, speed, accuracy trade-offs

Deliverable: Conceptual understanding of LLM mechanics and prompt design

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

10:30–11:30 | Local LLM Setup & Configuration (60 min)

Why local LLMs? Privacy, cost, latency benefits
Ollama vs LM Studio: comparison and selection
Step-by-step installation on Windows/Mac/Linux
Model selection: Mistral 7B, Llama 2, Neural Chat (performance vs resource trade-offs)
Running first inference: CLI commands and API endpoints
Performance tuning: GPU acceleration, quantization, batch processing
Troubleshooting common setup issues

Deliverable: Working local LLM instance running on participant laptops

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

12:00–13:00 | TypeScript + Vercel AI SDK Fundamentals (60 min)

TypeScript essentials for AI development: types, interfaces, async/await
Vercel AI SDK overview: architecture and core modules
API patterns: streaming vs non-streaming responses
Prompt engineering in code: templates, dynamic content, variable injection
Error handling and retry logic
Rate limiting and token management
Building a “Hello World” GenAI application

Deliverable: Hello World GenAI app with streaming responses

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

13:00–14:00 | Next.js + Shadcn UI Frontend Development (60 min)

Next.js 14+ fundamentals: App Router, API routes, server/client components
Project structure best practices for GenAI applications
Shadcn UI component library: setup, theming, customization
Building responsive UI: chat interface, input forms, output displays
State management: React hooks (useState, useContext) for conversation history
Real-time updates: streaming UI patterns with Suspense
Accessibility & UX patterns for AI applications

Deliverable: Functional frontend scaffold with chat UI and state management

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

14:00–14:45 | RAG & Tools Integration (45 min)

Retrieval-Augmented Generation (RAG) architecture: why it matters
Vector embeddings: concept, generation, storage
Open-source vector databases: Chroma, Milvus, Qdrant setup and usage
Document chunking strategies: semantic, sliding window, recursive
Retrieval pipeline: query embedding, similarity search, ranking
Tool calling & function integration: agent patterns, tool definitions
Building agentic workflows: decision trees, tool selection logic
Integration with local LLMs and Vercel AI SDK

Deliverable: RAG pipeline + tool integration working end-to-end

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

15:30 - 16:15 | Lab: Build & Deploy (45 min)

Hands-on: Participants integrate all components into one application
Local LLM backend + Next.js frontend + RAG pipeline + tools
Deployment options: Vercel, local Docker, cloud platforms
Testing and debugging: common issues and solutions
Performance optimization: response times, token usage
Code review and best practices feedback

Deliverable: Deployed GenAI application with all features working

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

16:15 - 17:30 | Live Demo Showcase & Awards (45 min)

12–15 best builds presented (3–4 min each)
Judging criteria: innovation, functionality, UI/UX, code quality
Live Q&A with presenters
Awards ceremony: Best Overall, Most Creative, Best RAG Implementation, etc.
Recognition and certificates

Deliverable: Recognition, community building, inspiration for participants

Featuring

Vassilis Karabalis
Head AI Development, Entersoftone Group

16:15 - 17:30 | Live Demo Showcase & Awards (45 min)

12–15 best builds presented (3–4 min each)
Judging criteria: innovation, functionality, UI/UX, code quality
Live Q&A with presenters
Awards ceremony: Best Overall, Most Creative, Best RAG Implementation, etc.
Recognition and certificates

Deliverable: Recognition, community building, inspiration for participants

17:30–18:00 | Networking & Closing Remarks (30 min)

Discord community setup for ongoing support
Resources for continued learning: documentation, GitHub repos, tutorials
Feedback collection and bootcamp survey
Closing remarks: next steps, community involvement

Attended by