AI Engineer · Autonomous Systems · LLM Pipelines
I build autonomous AI agents and production-grade automation pipelines that think, decide, and act — zero human in the loop. Currently building AutoApply at Zyngate. IEEE published. Patent approved.
Full-stack AI web application that automatically transcribes speech with word-level timestamps and burns perfectly synced captions into video frames. Built and deployed end-to-end — from zero to a live product with CI/CD, SEO, and a complete production pipeline.
Production agentic system that autonomously classifies support tickets, retrieves context via RAG over a vector knowledge base, reasons over routing logic, and executes decisions — end-to-end, no manual intervention. Reduced operational effort by ~30%.
Production-ready FastAPI application for Hindi speech recognition using NVIDIA NeMo's Conformer CTC model, optimized with ONNX for fast inference. Dual server modes, web interface, Docker containerized. Full builder ownership: idea → prototype → optimize → deploy.
AI-powered resume screening pipeline that extracts, parses, and scores candidate profiles against job requirements. Structured data extraction from unstructured documents feeding into downstream automation steps.
Deep learning-based classification of Areca Nut Yellow Leaf Disease using ResNet-50 CNN. Trained on multi-class leaf image dataset. Peer-reviewed and published at IEEE 2024 International Conference.
Image recognition system built with Python demonstrating computer vision pipeline from preprocessing to classification. Clean, modular codebase with well-documented inference pipeline.
Need an AI agent, automation pipeline, or LLM workflow? I ship fast, test everything, and write code you can maintain. Message me and I'll confirm feasibility before you commit.