Harshit Singh

I’m an AI/ML engineer with full-stack capabilities, focused on taking models from research to production. I work across the ML lifecycle—from data collection and PyTorch/TensorFlow training to deployment—using FastAPI and Python to build scalable, production-ready AI systems.

In my Full-Stack AI/ML Intern role at Caprae Capital Partners, I deepened my MLOps foundation through CI/CD, containerization, and end-to-end MLOps pipelines. I collaborated on cloud deployments with AWS and Docker, helped ensure reproducible production pipelines, and reduced deployment time by 35% while also integrating LLMs into system and UI.

I build practical solutions through hands-on projects, like an Audio CNN classifier with a serverless inference pipeline and interactive web interface, and an object detection system using OpenCV with Raspberry Pi for real-world lightning handling. I also developed an AI music generation platform that uses ACE-Step and Qwen 2.5 to generate music/lyrics from prompts and deployed scalable backend services with Modal and AWS S3.