Build with AI 2026: Gemma day I Speaker: Tarun Jain

Run On-Device AI Applications Without Internet Using Qdrant Edge and Google LiteRT | Tarun Jain | Build with AI 2026 Can you build a complete vector search and AI reasoning pipeline on embedded devices without sending a single byte over the internet? In this technical session from GDG Ahlen's virtual Build with AI 2026: Gemma Day, Tarun Jain explores how to architect fully offline, on-device AI applications. Learn how to combine Qdrant Edge (an in-process vector search engine designed for resource-constrained hardware) with Google’s LiteRT to run Gemma 4 locally, creating self-contained retrieval and inference pipelines at the edge. 📌 What You’ll Learn in This Video: • Getting started with Qdrant Edge for embedded, in-process vector search • Deploying and running Gemma 4 locally with Google LiteRT on resource-constrained devices • Building end-to-end on-device Retrieval-Augmented Generation (RAG) pipelines • Designing resilient, zero-connectivity AI systems with absolute data privacy and ultra-low latency 👤 Speaker: Tarun Jain — Software & AI Engineer specializing in edge computing, vector databases, and on-device machine learning architectures. 🔗 Resources & Links: • you can refer to github.com/unslothai/unsloth • for hands-on notebooks: github.com/unslothai/notebooks • github.com/lucifertrj/offline-on-device-adi • docs.google.com/presentation/d/1YV0S_A-DIOozD30fyAaOcxVH5Rf65pMlz3JtIGlw-aE/edit?usp=sharing Recorded during the virtual Build with AI 2026: Gemma Day, organized by Google Developer Groups (GDG Ahlen). #EdgeAI #Qdrant #LiteRT #Gemma #OnDeviceAI #VectorSearch #RAG #GDGAhlen #BuildWithAI #MachineLearning