Shubhrant Mishra

I’m a Backend Engineer focused on building production systems for an AI-based video generation platform and vendor/sales management software. I bring strong expertise in Node.js, TypeScript, Python, and media-rendering pipelines that run reliably on cloud functions and queue-driven architectures.

In my current role, I develop backend services in Node.js for distributed video generation using GCP Cloud Functions and Cloud Tasks. I’ve driven key reliability improvements by migrating text-template rendering from FFmpeg per-character filter chains to a Canvas-based 2-pass model—eliminating cold-start font crashes and expression-parser failures while restoring broken animations.

I also built and delivered a SmartCrop engine that combines keyword heuristics with an ONNX (YuNet) face-detection model to keep subjects centered during 16:9 rendering. Across a monolith-to-microservices transition, I transferred render templates into independently scaled services and built backend APIs backed by MySQL, deploying on AWS Lambda and GCP Cloud Functions with a Terraform-based deployment architecture. I was recognized as “Rising Star of the Year” within 4 months of joining for this impact.