IoT & ML Pipeline: Architecture & Live Demonstration

Join us for an immersive architectural walkthrough and live technical showcase where we break down how to build an end-to-end IoT machine learning pipeline on AWS using an ESP32 microcontroller and an MPU6050 motion sensor. (Similar Tech Used in a Tesla/Waymo). Room (INV1 - 201) In this session, we will live-demonstrate how physical gestures (idle, shake, circle, and tilt) stream directly from hardware into the cloud, get processed through a serverless data lake, and are used to train and deploy a compact machine learning model right back onto the edge device. What You'll Learn: Device Identity & Ingestion: See how the ESP32 securely connects to AWS IoT Core using MQTT for real-time telemetry streaming. Serverless Data Lake Architecture: Follow the data flow as IoT Rules route batches into Amazon Kinesis Firehose and Amazon S3. Serverless Analytics: Watch raw JSON payloads queried and explored instantly using Amazon Athena and SQL. Edge Machine Learning: Understand how generated gesture datasets are leveraged to train lightweight models for on-device inference using TFLite Micro. Real-World Scaling: Get a sneak peek at a live drone telemetry demo showing how this exact pipeline scales to complex edge systems. Agenda: 3:00 PM – 3:15 PM: Check-in, grab a seat, and settling in 3:15 PM – 3:45 PM: Architecture deep-dive: The 5 core AWS services and fundamentals 3:45 PM – 4:30 PM: Live demonstration: Data ingestion, S3 storage, and Athena SQL exploration 4:30 PM – 4:50 PM: Model training workflow & deploying back to the ESP32 4:50 PM – 5:00 PM: Scale demo (drone telemetry) & Q&A What to Bring: Your laptop or phone to follow along with shared dashboards and architecture diagrams. Your questions on cloud architecture, IoT, and edge machine learning!