Avinash Chadalavada

Project: IoT-enabled Safety Smart Watch — Data Platform, ML Pipelines & PTW Architected end-to-end data platform for IoT wearable streams: ingestion → real-time PySpark/Kafka processing → AWS cloud storage → Databricks analytics layer processing millions of events/day →Designed and maintained robust ETL pipelines in Python for high-volume telemetry data, applying data transformation, cleansing, and Data Quality Rules to ensure reporting-ready datasets →Built real-time ML fatigue detection pipeline on wearable sensor streams; surfaced risk scores and heatmaps via Databricks One dashboards for live monitoring by safety officers and ops managers →Led development of complex SQL queries and scalable database load processes on TimescaleDB and PostgreSQL ensuring optimal data storage and retrieval at scale →Implemented Kafka + Zookeeper streaming infrastructure processing millions of device events/day with low latency on AWS; designed automated SNS-based alert and escalation workflows →Drove AI/Data strategy: defined platform roadmap, partnered with Data Science to productionise ML models into live Databricks pipelines; led 20-member global team with agile delivery and full SDLC practices →Gen AI POC: LangChain-orchestrated RAG pipeline over Databricks Genie enabling natural-language queries over live IoT telemetry — validated as a replacement for manual report generation

Led and managed a 13-member globally distributed Cloud, Data Engineering & Data Science team; drove agile delivery, mentorship, performance cadence, and cross-functional collaboration Architected end-to-end data lifecycle platform — real-time telemetry ingestion via Kafka, batch processing with PySpark on Databricks — processing millions of IoT events/day Designed and maintained ETL pipelines in Python for integrating high-volume datasets to match reporting requirements; implemented data transformation, cleansing, and Data Quality Rules frameworks to enforce data integrity, completeness, and accuracy Implemented Kafka + Databricks stream and batch processing pipeline for warehouse automation; improved query response times by 40% through caching, multithreading, and pipeline optimisation Developed complex SQL queries and scalable database load processes on PostgreSQL and Cassandra; leveraged MongoDB for document-based NoSQL workloads — managing schema design, indexing strategies, and high-throughput read/write patterns Built and maintained Power BI dashboards surfacing warehouse KPIs, occupancy trends, and churn risk scores to business stakeholders — integrated directly with Databricks and PostgreSQL for self-serve BI reporting Designed high-performance APIs handling millions of IoT data points with low latency on AWS and Azure; applied microservices and event-driven architecture patterns with asynchronous communication Built production ML pipeline (Gradient Boosting & Prophet) on Databricks/Spark to predict customer churn at scale, enabling proactive data-driven retention strategies Deployed computer-vision model into real-time data pipeline for motion-based staff optimisation; Gen AI POC using Azure OpenAI (GPT-4) + LangChain for natural-language summaries of environmental data Key founding engineer (team of 5 → 13); engineering culture; applied full SDLC best practices including source control, build processes, automated QA, and CI/CD

• Led the end-to-end IoT system architecture, integrating edge computing, real-time telemetry, and cloud services with downstream data engineering and data science workflows. • Founding team member (initial 5 members) – played a pivotal role in early product development, team expansion, and scaling of processes and systems. • Designed and implemented AI/ML pipelines, leveraging existing models such as Prophet and residual correction techniques to improve forecasting accuracy. • Built and optimized data engineering pipelines using Databricks, enabling efficient data processing and integration for advanced analytics. • Developed scalable APIs and web architectures capable of handling millions of requests, ensuring system resilience and high availability. • Engineered IoT-enabled smart storage solutions for warehouse management using mesh networking technologies. • Delivered AI-powered front-end automation tools with React and Postman, improving user workflows and operational efficiency. • Drove cross-functional collaboration with designers, product managers, and engineers, resulting in improved frontend performance, accessibility, and reduced load times. • Optimized embedded Linux firmware (OpenWRT, UCI, UBus) for seamless connectivity, stability, and scalability across IoT devices

Was Thingworx’s lead engineer for 3 projects. Worked on architectural solution from scratch. Worked in complex integration to AWS Athena to Thingworx’s

Worked majorly in 1. AWS IoT- Designed and developed the whole Command Center Using AWS services. 2. PTC ThingWorx - Worked with Client on there Digital journey to Smart Factory using PTC Thingworx.

Worked Majorly in ThingWorx IoT - Worked on Internal Asset Creation in PTC ThingWorx Azure IoT - Worked as a developer hand in hand with client to build Device Management solution.

Worked Majorly in PTC ThingWorx - MES Integration with ThingWorx MES - GE proficy MES IBM Watson - Certified on Watson IoT. GE Predix - Did PoC on Predix and was runner up in GE prefix hackathon..

* Worked in Cryogenic Engine ignition timing system which was used in GSLV D-5 * Trained in Vehicle Assembly Station and Material Handling equipment of the space vehicles.