Skip to content
View harshit-862000's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report harshit-862000

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
harshit-862000/README.md

Typing SVG

About Me

I am an AI Engineer with a focus on building production-ready Large Language Model (LLM) and Retrieval-Augmented Generation (RAG) systems. I enjoy working at the intersection of applied research and engineering — turning complex models into reliable, scalable applications.

  • Currently building RAG pipelines for intelligent document understanding using Mistral, LlamaIndex, and vector databases
  • Deepening expertise in LangChain & LlamaIndex — agentic workflows, memory management, and advanced retrieval strategies
  • Open to collaborating on open-source AI/ML projects, especially in NLP and LLM application development
  • Areas of interest: MLOps, LLM deployment, and scalable AI infrastructure
  • Reach me at: LinkedIn · Email

Tech Stack

Languages & Core

Python SQL

AI / ML Frameworks

PyTorch TensorFlow scikit-learn HuggingFace LangChain

Data & Experimentation

Pandas NumPy Jupyter Google Colab MLflow

Tools & Platforms

Git GitHub Docker FastAPI VS Code

Featured Projects

DocuMind — RAG-Powered PDF Q&A with Mistral 7B

A production-oriented Retrieval-Augmented Generation pipeline for natural language question answering over PDF documents. Built with Mistral-7B-Instruct (8-bit quantized), HuggingFace sentence-transformers for semantic embeddings, and LlamaIndex for vector indexing and query orchestration.

Python LlamaIndex Mistral 7B HuggingFace BitsAndBytes

Connect

LinkedIn GitHub Gmail

Popular repositories Loading

  1. Heart-Disease---Classifications-Machine-Learning- Heart-Disease---Classifications-Machine-Learning- Public

    Predicting heart disease risk from clinical data using 6 ML classifiers — Logistic Regression, KNN, SVM, Naive Bayes, Decision Tree, and Random Forest — with full EDA, preprocessing, and model comp…

    Jupyter Notebook

  2. Finance-Tracker Finance-Tracker Public

    The Finance Tracker is a dynamic web application built using Django, Python, HTML, CSS, and JavaScript designed to help users efficiently manage their personal finances. It allows users to securely…

    CSS

  3. Url-shortner Url-shortner Public

    About a lightweight url shortener (with frontend + api) in next/react ready to be deployed out of the box with vercel. demo: https://lite.fyi

    JavaScript

  4. Leveraging-Machine-Learning-in-Distributed-Systems Leveraging-Machine-Learning-in-Distributed-Systems Public

    Built a distributed kNN system using Python and the Ray framework to scale machine learning workloads. Parallelized computation across 10+ workers, reducing execution time by ~40% and improving thr…

  5. CRUD-Application-using-FastAPI CRUD-Application-using-FastAPI Public

    A cloud-ready Task Management REST API built with FastAPI, MongoDB (Motor), and GitHub Issues integration — containerized with Docker and deployable on Minikube with HPA support.

    Python

  6. Text-Summarizer Text-Summarizer Public

    Abstractive text summarization system that fine-tunes Google's PEGASUS sequence-to-sequence transformer on the BBC XSum dataset. Uses the Hugging Face Trainer API with gradient accumulation for mem…

    Jupyter Notebook