KrishiMitra AI
Boosts farmer efficiency by diagnosing crop diseases from a photo in seconds, scraping live market prices, and providing local weather forecasts, reducing reliance on manual expert consultation.
A Seer of systems and a Fool for elegant code. I specialize in backend architecture and applied AI, advancing projects from Sequence 9 to production-ready reality.
Shivam Rathod
System Architect • AI/ML Developer
My work is about divination in a technical sense: perceiving the final, stable structure within a complex set of requirements. I chart a clear pathway from abstract concepts to dependable backend systems and applied AI/ML solutions. Whether architecting a bilingual AI assistant or detecting malicious code, my focus is always on a logical sequence of development that results in a clean, measurable, and reliable product.
A breakdown of my technical skills. The interactive constellation offers a more esoteric view of their connections.
Selected work. Filter by focus.
Boosts farmer efficiency by diagnosing crop diseases from a photo in seconds, scraping live market prices, and providing local weather forecasts, reducing reliance on manual expert consultation.
A collection of hands-on agents using local LLMs (Ollama) to perform tasks like web scraping, voice assistance, and PDF Q&A, showcasing modern agentic workflows.
A NASA Space Apps project that detects exoplanets from raw light curve data with ~90% accuracy. Its physics-driven AI performs independent classification, unlike systems that rely on existing labels.
A multilingual chatbot for the Smart India Hackathon that reduces administrative workload by over 40% by giving students instant, trusted answers from official college documents via a RAG pipeline.
A CyberShield Hackathon project that detects malicious Android banking APKs with 95% accuracy. It provides explainable AI insights, showing users exactly which permissions make an app dangerous.
A complete, production-ready e-commerce platform built with Django. Demonstrates secure authentication, session-based cart management, and a deployment-ready setup using Gunicorn and WhiteNoise.
Developed the FastAPI backend for a field data verification platform, enabling automated photo validation and job tracking via a bilingual WhatsApp interface.
Contributed to a one-month intensive program focused on rapid development and creating efficient, reusable Python code to solve a variety of challenges.