Anukul User

I’m a software developer focused on turning research-grade AI into scalable, real-world applications. I’ve built end-to-end platforms and shipped models that perform reliably under real usage constraints.

As CTO at CHEM0 Corp Limited, I built a scalable full-stack chemical reaction prediction platform supporting 1,000+ concurrent beta users across 3 academic institutions. I also designed and trained a Graph Neural Network (GNN) model, reaching 92.84% accuracy on 10,000+ molecular structures.

During my DRDO, Ministry of Defence internship, I developed and fine-tuned a YOLO-based computer vision system for human detection using thermal imagery. I trained and benchmarked models across 20+ scenarios, achieving 94.36% detection accuracy with optimized real-time inference.

I enjoy building product-minded AI systems, like AgentFlow—a visual AI agent builder with DAG workflows, real-time preview, REST API persistence, and exportable JavaScript snippets—and BazARmash, an on-chain NFT voting platform leveraging an AO smart process and Arweave wallet authentication. I’ve also earned an IEEE publication (2025) and multiple hackathon awards, alongside certifications in Machine Learning, Computer Vision, and NLP.