Medhansh Jain

I’m a software engineer intern with hands-on experience building production backend systems and improving end-to-end product workflows. At my SDE internship, I redesigned campaign template creation using a Bulk CSV Upload pipeline, added centralized validation and history tracking, and delivered faster batch processing with full traceability.

I also optimized reliability and operating costs in distributed processing by re-tuning Apache Flink checkpointing and AWS S3 filesystem configuration, cutting S3 costs by approximately $4,000/month while preserving stable execution and reliable failure recovery. I fixed a real user impact issue by implementing IST/UTC timezone-aware day-boundary support for daily reward aggregation so rewards align correctly with each user’s local day.

Alongside backend work, I pursue applied ML: I built an overspeeding vehicle detection system with speed estimation and automatic license plate recognition, and improved object detection accuracy from 86.5% to 94.7% using a two-stage transfer learning pipeline with YOLOv8n. My projects—from a Spring Boot finance API to an NLP-driven hybrid recommender—reflect how I combine strong engineering fundamentals with data-driven outcomes.