Build with AI with Google, Tata 1mg

Important Notes: RSVP on the AICamp event website OR luma is required for admission.

Description:
Welcome to the AI meetup in Delhi, in collaboration with Google, Tata 1mg and Snowflake. Join us for deep dive tech talks on AI, GenAI, LLMs and machine learning, hands-on experiences on code labs, workshops, and networking with speakers and fellow developers.

Agenda:
* 10:00am~10:30am: Checkin and networking
* 10:30am~1:00pm: Tech talks and Q&A
* 1:00pm: Open discussion and Mixer

Tech Talk: ​How AI Learns to Decide Which Store Ships Your Order: Deep RL and Constraint Optimization for Store Assignment
Speaker: Arkya Bagchi
Abstract: This workshop explores how AI can optimise store assignment in complex e-commerce supply chains. We’ll introduce an attention-based model that learns existing fulfilment policies through supervised learning and is then fine-tuned using reinforcement learning with an Advantage Actor-Critic (A2C) framework. The session will cover how delivery time, fulfilment cost, inventory availability, and split shipments can be jointly considered through a simulation-based reward. We’ll also discuss the practical challenges of deploying such models in large-scale, real-world supply chain systems and insights from experiments on Tata 1mg’s e-pharmacy network.

Tech Talk: ​Building a Clinical Digital Twin — Learning Patient State from Sparse Health Records
Speaker: Anshul Patel
Abstract: What would it take to build a "digital twin" of a patient — a compact representation that captures how someone's health is evolving over time, learned directly from messy, irregular clinical records? This talk introduces a fresh take on the Digital Twin project: learning a reusable longitudinal patient-state representation from sparse lab panels and reports. We'll look at how the design cleanly separates a patient's underlying clinical state from the interventions applied to them and their static context. We'll walk through the practical machinery — encoders that respect irregular time gaps, treating missing data as genuinely unavailable rather than "normal," and building incrementally so every component earns its place. We'll close with where this is headed: from representing patient state to explanation and forecasting.

Tech Talk: ​ Run On-device AI Applications without the internet using Qdrant Edge and Google LiteRT
Speaker: Tarun Jain, Founding Engineer, Kaivid Labs
Abstract: Running AI applications without an internet connection is increasingly important for privacy and resource constrained environments. In this talk, we will cover how to build fully on device retrieval and inference pipelines using Qdrant Edge and Google LiteRT, with a Streamlit UI for the demo. We will also look at performance and the tradeoffs involved in running on-device application.

Tech Talk: ​ Building an Autonomous Financial Audit Agent Team

Speaker: Sumit Kumar, Cloud Solution Architect, 66degrees
Abstract: Discover how to build a production-ready autonomous financial audit system using Google's Agent Development Kit (ADK) and Google Cloud. In this hands-on session, you'll learn how a team of AI agents can collaborate to reconcile financial records, process invoices from Cloud Storage, validate ERP data in BigQuery, and generate audit reports using Vertex AI. We'll also cover key design patterns for building secure, scalable, and enterprise-grade multi-agent AI applications.

Sponsors:
We are actively seeking sponsors to support AI developers community. Whether it is by offering venue spaces, providing food, or cash sponsorship. Sponsors will not only speak at the meetups, receive prominent recognition, but also gain exposure to our extensive membership base of 10,000+ AI developers in Delhi and 500K+ worldwide.