About
As a Data Science and MLOps enthusiast, I specialize in designing, building, and deploying end-to-end ML pipelines using Python, MLflow, and Docker. My experience spans CI/CD automation, AWS deployment, and agentic AI workflows using LLMs. I'm particularly focused on applying data-driven insights and AI automation in finance and analytics to enhance decision-making and efficiency.
Work Experience
Skills
Check out my latest work
I've worked on a variety of projects, from machine learning systems to agentic AI workflows. Here are a few of my favorites.
MLOPS-LLMOPS: End-to-End Crop Advisor & Agentic LLM Project
Built a full-scale MLOps + LLMops pipeline for crop analysis: ingest data, feature-engineer NPK & environmental variables, train models (Decision Tree / RF / XGBoost), track experiments with MLflow, version data with DVC, deploy via FastAPI + Streamlit, containerize via Docker, and orchestrate on AWS EKS/ECR. Integrated an LLM agent to provide natural language agronomic advice.
Agentic RAG System for Game Simulation Agents
Developed an agentic RAG system to simulate intelligent agents that reason and interact in a game environment. Integrated LangGraph, Groq LLM, and FastAPI backend with a JavaScript-based frontend. Implemented vector retrieval with MongoDB and built observability using Opik for LLM evaluation.
Sentiment Classification System (IMDB Dataset)
Built a sentiment analysis model to classify IMDB movie reviews. Implemented text preprocessing using NLTK, trained multiple models (Logistic Regression, SVM), and deployed via Flask API on AWS EKS. Used DVC & MLflow for versioning and monitoring with Prometheus & Grafana.
Certifications & Achievements
These certifications reflect my continuous learning and growth in Data Science, MLOps, and AI.