Naman Bansal

About Naman Bansal

I’m a Data Scientist at Wiliot, where I design and deploy machine learning systems that…

Experience

1. Data Scientist

   Wiliot

   Feb 2024 — Present · San Diego, CA, US

   Trained and deployed a production-grade machine learning model for temporal event detection for Walmart’s IoT infrastructure, enabling in-store item movement tracking with 99.5% accuracy.• Engineered scalable real-time machine learning pipelines using PySpark and Databricks to process data from over 9M IoT sensors with less than 2 second latency enabling instant analytics.• Prototyped LLM-based tools using LangChain and OpenAI APIs to enable natural language querying of sensor data. Implemented a RAG memory-layer with vector similarity search to get the most relevant context.• Executed end-to-end feature engineering on multivariate time series data using NumPy, Pandas, and SQL— partnered with business teams to validate data patterns via hypothesis testing and statistical inference.• Designed visualizations and analytics dashboards to monitor model performance, drifts and coverage — enabling stakeholders to make data-driven decisions across operations and engineering.