01 — About
Building from
first principles.
I'm Naman Bansal — a software engineer and AI builder based in San Francisco. Currently SDE at Wiliot, where I've spent two years building ML systems that process data from over 9 million IoT sensors in real time.
I grew up in Akola, Maharashtra, always drawn to technology and building. I don't follow playbooks — I go back to first principles, understand the problem at its core, and build something real.
Full story →02 — Experience
Where I've worked.
Wiliot← nowFeb 2026 — Present
Software Development Engineer
- –Leading software development on IoT infrastructure powering enterprise-scale deployments.
- –Previously built ML systems achieving 99.5% accuracy on temporal event detection for Walmart.
WiliotFeb 2024 — Feb 2026
Data Scientist
- –Engineered real-time ML pipelines on PySpark/Databricks processing 9M+ IoT sensors at <2s latency.
- –Prototyped LLM tools with LangChain + RAG for natural language querying of sensor data.
iMochaAug 2021 — Aug 2022
Software Developer
- –Built multilingual code generation tool supporting 13+ languages, reducing generation time by 70%.
- –Automation framework scaling test coverage across 1,000+ problems, cutting QA cycles by 80%.
03 — Projects
Things I've built.
Data2dialog
LangChain · OpenAI · Django · React · AWSProduction Text-to-SQL engine. Natural language → executable queries across CSV, Excel, and SQL databases.
Quizzx
Python · FAISS · NLP · JavaScriptAI-driven quiz generation from documents using semantic similarity search and FAISS vector indexing.
Vusion Edge Model
Computer Vision · Edge AI · PythonEdge-optimized vision model for real-time retail shelf recognition and inventory tracking.