Dipanshu Yadav

Lead a team of 5 engineers designing and delivering cloud-native dbt transformation pipelines on AWS, powering scalable and high-performance enterprise analytics. Architected end-to-end ETL workflows producing Analytical Ready Datasets and Model Ready Datasets for credit risk modelling.

Architected an ML-based risk assessment framework for the Roadzen (insurtech) partnership, improving fraud detection accuracy by 35% and reducing claims processing time by 25%.

Engineered a driver risk scoring algorithm integrating vehicle telemetry, weather APIs, traffic conditions, and computer-vision analysis of dashcam footage, reducing insurance risk assessment errors by 30% across 1,000+ commercial fleet vehicles. Designed a carbon emissions monitoring system using supervised learning to predict Scope 1 and Scope 2 emissions.

Developed a hybrid recommender engine addressing the long-tail effect using collaborative and content-based filtering, applying diversity-promoting re-ranking and popularity-bias regularisation to improve recommendation diversity by 25% and user satisfaction by 15% across a 10,000+ user validation cohort.

Grade: CGPA: 3.28 / 4.0

Activities and societies: Relevant Coursework: Linear Algebra, Probability & Statistics, Machine Learning, Pattern and Anomaly Detection, Advanced Data Structures, Design and Analysis of Algorithms, Design Thinking, Venture Ideation

Bachelor of Technology in Computer Science and Engineering from the University of Petroleum and Energy Studies, Dehradun, India, from 2019 to 2023.