Dynamic Machine Learning and AI Engineer skilled in developing and deploying production-ready ML and GenAI systems — including RAG pipelines, intelligent agents, and MLOps workflows. Experienced with TensorFlow, YOLOv8, JAX, LangChain, and LLM orchestration. Passionate about building scalable, secure, and automated solutions that enhance product intelligence and reduce manual effort.
Dynamic Machine Learning and AI Engineer skilled in developing and deploying production-ready ML and GenAI systems — including RAG pipelines, intelligent agents, and MLOps workflows. Experienced with TensorFlow, YOLOv8, JAX, LangChain, and LLM orchestration. Passionate about building scalable, secure, and automated solutions that enhance product intelligence and reduce manual effort.

I’m a B.Tech graduate in Estate Management turned Machine Learning and AI Engineer, passionate about transforming data into intelligent systems.
I specialize in building and deploying production-ready ML and GenAI solutions — from RAG assistants and intelligent agents to MLOps pipelines.
My work lives at the intersection of data, automation, and innovation, creating AI-driven tools that enhance decision-making and drive measurable impact.

Designed and developed a LLM powered multi tenant customer service solution using Gemini, Langchain, Python, superbase, pinecone, cloudflare R2 buckets that allows businesses to simply upload their documents and serve their customers a customer support solution instantly.

Built a FastAPI + LangChain app using Gemini embeddings to semantically rank CVs against job descriptions, generating CSV/PDF shortlists. Deployed on DigitalOcean with CI/CD and unit tests.

Developed an LLM-powered tool to ingest research materials, auto-generate literature-review tables, and export to CSV/PDF, streamlining academic workflows.

Ported ML models to JAX/Flax for performance optimization, demonstrating adaptability to modern frameworks.
Designed and deployed a mentorship matching recommender system and a secure RAG Knowledge Assistant using LangChain, Pinecone, AWS S3, and PostgreSQL. Implemented RBAC/ABAC access controls, optimized response time and retrieval precision, and integrated monitoring (Prometheus/Grafana) with runbooks for deployment, incident management, and data retention.
Co-developed Hey Nova AI, an Azure OpenAI–powered educational agent with live internet access, built using FastAPI, Azure AI Search, embeddings, and Redis to support 200+ concurrent users. Implemented automated summarization and question generation for enhanced learning engagement. Also created an Azure AI Foundry–based academic recommender system for personalized insights and designed a YOLOv8/InceptionResNet security prototype (90%+ accuracy) integrated with FAISS for event retrieval and loss reduction.
Developed a Zero-Shot Content Moderation Bot using Hugging Face LLMs (BERT) that reduced manual review workload by 40%. Implemented retraining pipelines and evaluation scripts for continuous model optimization.