Niranjan Poudel

I’m a Data Scientist who turns complex travel, customer, and geo-spatial problems into practical ML and analytics products. I like building end-to-end solutions—ETL automation, modeling, and production-ready delivery—so teams can move faster with trustworthy insights.

In my most recent role as a Sr. Customer Insights Consultant, I automated ETL from manual code runs to BigQuery scheduled procedures using airflow (Google Composer) and GitHub. I created customer-facing dashboards with Looker (Google Cloud) and React (JS), and I developed AI Chat bots using RAG models and Gemini AI for both looker and non-looker dashboards.

Before that, at Overair Inc., I modeled and solved network optimization using mixed integer methods with stochastic travel demand—then layered deep learning for induced demand. I also built unsupervised machine learning (geo-spatial time series clustering) for weather regionalization and delivered web dashboards using Leaflet/Mapbox, plus live performance tracking.

I’ve also worked across research and applied engineering: conducting stated preference surveys and USDOT-funded studies on bicyclist comfort, safety, and perceptions, and analyzing traffic safety concerns. Earlier as a Civil Engineer in Nepal, I supported data analysis, development plans, and inspections—habits that still shape how I approach rigorous, decision-ready analysis.