Build Your Own Offline AI Assistant with EmbeddingGemma 2

A hands-on workshop for developers: Build a local RAG pipeline with Google's latest embedding model.

90 minutes | Intermediate | €5 | Limited to 30 participants

Your documents. Your AI. Your hardware.
What if you could build an AI assistant that understands your documents, answers your questions, and cites its sources, without sending anything to the cloud?

In this 90-minute hands-on workshop, you'll explore Google's newly released EmbeddingGemma 2 and build a local Retrieval-Augmented Generation (RAG) pipeline using Python.

Starting with a folder of PDFs, we'll turn documents into embeddings, implement semantic search, and connect everything to a local LLM to generate answers grounded in your own data.

No cloud APIs. No per-query fees. Your documents stay on your machine.

What you'll build and learn

Build a local RAG pipeline: Process PDFs, generate embeddings with EmbeddingGemma 2, and retrieve relevant content using vector similarity search.

Connect a local LLM: Ask questions in natural language and generate answers based on your documents.

Add citations and guardrails: Reference original page numbers, verify retrieved evidence, and handle questions the documents cannot answer.

Run everything locally: Understand how to build privacy-first AI applications without external API dependencies.

Inspired by Agora Library, we'll walk through the complete pipeline from document ingestion to source-grounded answers.

You'll leave with a practical blueprint you can adapt to your own projects.

Meet Your Instructor
Francisco Cordoba

CTO at Snapscribe, Marie Curie Research Fellow at the Institute for the Future, and author of Beat the Robots.

Francisco is also an MIT Innovator and Bank of Cyprus Hackathon winner, bringing research and hands-on technology experience to this workshop.

Who should attend?
This is an intermediate-level technical workshop, designed for:

Backend and full-stack developers interested in RAG, embeddings, and vector search.

AI/ML engineers and tech leads exploring local-first AI architectures.

Computer Science students and researchers interested in private document search and local LLMs.

No previous RAG or vector database experience required. You should be comfortable with Python, the terminal, and basic API or ML concepts.

Before the workshop: Prepare your laptop
To make the most of our 90 minutes, please arrive with your development environment ready.

Required setup:

Laptop with Python 3.10+ installed

At least 8 GB of free disk space

A code editor (VS Code, Cursor, Antigravity IDE, or similar)

Working terminal and Python environment

Ollama installed

Required Hugging Face libraries and EmbeddingGemma 2 model weights downloaded

Make sure your laptop is charged and capable of running local models.

Important: The workshop focuses on building, not installing tools. Please complete the setup before attending.

Reserve your spot
Only 30 seats available.

Entry: €5 per participant

Bring your laptop, get hands-on with EmbeddingGemma 2, and discover how to build AI applications that run entirely on your own hardware.

[Reserve Your Spot - €5]

Agenda

---

Hosted By

Maria Stylianou, GDG Organiser

I'm an engineering leader, passionate about building stuff, experimenting and failing fast. My expertise lies in up-skilling developers and scaling tech organizations. I advocate for increasing diversity in tech, and creating truly inclusive and fun environments.

I consult tech organizations to increase their efficiency, with trainings, team facilitations and individual tech coaching.

Hit me up for a challenge, advice or coffee!

Andreas Lefkatis, GDG Organizer

I am a Greek Cypriot Software Developer passionate about Web, DevOps and Mobile Application Engineering.
My interests lie in how cloud based infrastructures and mobile devices could revolutionise ubiquitous computing and today's tech world.
GDG Cyprus Lead and HackCyprus Organiser

Marilyn Haber, Software Developer | WTM Ambassador

Jawaad Farooq, Software Developer | Product

Maria Kruglova, program manager

Hey, there
I'm Masha, and I led GDG Kaliningrad, a community from Russia's smallest region, for 10 years. Now I live in Cyprus.
I co-founded Podlodka Crew, an online tech conference, and DevCrowd, a research project on the tech industry.
I am also the Ambassador of the WomenTechmakers chapter in our town.

---

Partner

UCY Computer Science Department (https://www.cs.ucy.ac.cy)

---

Complete your event RSVP here: https://gdg.community.dev/events/details/google-gdg-cyprus-presents-build-your-own-offline-ai-assistant-with-embeddinggemma-2/.