Skip to content
View ShubhaMahobia's full-sized avatar
🏠
Working from home
🏠
Working from home
  • Bhopal , Madhya Pradesh ,India
  • 23:00 (UTC +05:30)

Block or report ShubhaMahobia

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
ShubhaMahobia/README.md

🧠 About Me

class ShubhamMahobia:
    role        = "AI Engineer"
    location    = "Bhopal, Madhya Pradesh 🇮🇳"
    experience  = "2.5+ years in AI"
    focus       = ["RAG Systems", "LLM Pipelines", "Agentic AI", "Evaluation-Driven Optimization"]

    achievements = {
        "document_search_speedup" : "75% reduction in search time",
        "answer_accuracy"         : "92% with production RAG",
        "daily_docs_processed"    : "10,000+ enterprise documents",
        "screening_time_reduction": "30 min → 8 min per candidate",
    }

    currently_building = "Production-grade AI systems that scale & evaluate well"
    open_to            = "Full-time AI / ML Engineering roles"

📊 Impact at a Glance

🎯 Metric 📈 Result
Document Search Time ⬇️ 75% faster
RAG Answer Accuracy 92%
Docs Processed Daily 📄 10,000+
EU AI Act Classifier 🏷️ 85% accuracy
Interview Screening 30 min → 8 min
OCR Digitization 🖊️ 1,200+ handwritten docs @ 82%
Expense Extraction 💸 94% accuracy across 500+ transactions

🚀 Featured Projects

🔀 Agentic RAG System — Adaptive Retrieval with Routing Intelligence

LangGraph · FAISS · OpenAI API

  • Intelligent RAG router that dynamically picks: FAISS vector search → Web search → Direct LLM based on query type
  • +40% answer relevance over baseline RAG
  • RAGAS evaluation: Precision 0.89 | Recall 0.86 | Accuracy 0.91 | Faithfulness 0.87
  • Tested over 200+ queries across technical docs and general knowledge

💸 SpendSense — Expense Intelligence Backend with MCP

Python · LangGraph · MCP · FastAPI

  • Extracts & categorizes expenses from SMS/email using GPT-4 + Model Context Protocol
  • 94% extraction accuracy across 15 categories on 500+ test transactions
  • Multi-agent workflow with natural language query capabilities

🤖 Resumai — AI-Powered Resume Analyzer

RAG · FastAPI · Semantic Search

  • Dual-mode AI tool for both job seekers (resume feedback) and recruiters (candidate ranking)
  • Built on production RAG pipeline with semantic retrieval and structured output

📝 AI Blog Generation System — Multi-Agent Orchestrator

LangGraph · GPT-4 · Web Search

  • Planner agent decomposes topics and dynamically spawns 2–8 parallel worker agents
  • Successfully generated 50+ technical blogs with coordinated agent context sharing

🧪 AI Interview Automation Platform (@ Control Shift)

Gemini LLM · RAG · FAISS · sentence-transformers

  • Automated technical screenings for 100+ candidates
  • Semantic candidate–job matching at 78% ranking accuracy on 300+ profiles

🛠️ Tech Stack

LLM & GenAI

LangChain LangGraph OpenAI Gemini Claude RAG RAGAS LoRA

Vector Databases

Pinecone FAISS Weaviate ChromaDB

ML Frameworks

PyTorch HuggingFace scikit-learn TensorFlow

Backend & Cloud

Python FastAPI Docker AWS GCP Azure GitHub Actions


💼 Experience

🏢  RMgX LLP         →  AI Engineer Intern        [ Sept 2025 – Mar 2026 ]  📍 Gurugram
🏢  Control Shift    →  AI Intern                 [ Mar 2025 – May 2025  ]  📍 Gurugram
🏢  Mitt Arv         →  Software Engineer Intern  [ Jan 2023 – Nov 2024  ]  📍 Hyderabad

🎓 Education & Certifications

B.Tech in Computer Science — SAGE University Bhopal | CGPA: 8.86 / 10 | 2021–2025

📜 Certification 🏛️ Issuer
Model Context Protocol (MCP) Anthropic
Deep Learning IIT Ropar (NPTEL)
Generative AI with LangChain & HuggingFace Udemy
Supervised Machine Learning Coursera
Python for Data Science, AI & Development IBM

💬 What People Say

"Shubham showed remarkable dedication, a strong grasp of ML fundamentals and the ability to independently explore complex technologies."Rahul Singh, CEO, Control Shift

"He is very sincere, hardworking and a competent student… takes initiative on projects assigned to him."Vishal Mehta, Founder, Mitt Arv


🤝 Open to Full-Time AI / ML Roles

Let's Talk View Portfolio View Resume

Pinned Loading

  1. ChatTube-Know-about-your-video ChatTube-Know-about-your-video Public

    A comprehensive AI-powered application that enables intelligent conversations with YouTube video content using Retrieval-Augmented Generation (RAG). Built with FastAPI backend, Streamlit frontend, …

    Jupyter Notebook

  2. DVC-Pipeline-AWS DVC-Pipeline-AWS Public

    A comprehensive Machine Learning pipeline for SMS Spam Detection using DVC (Data Version Control) for reproducible ML workflows and DVC Live for experiment tracking. This project demonstrates best …

    Jupyter Notebook

  3. RAG-Based-Custom-Medical-Dataset-Chatbot RAG-Based-Custom-Medical-Dataset-Chatbot Public

    A GenAI-powered chatbot that leverages Retrieval-Augmented Generation (RAG) to answer medical questions using information from PDF documents. Built with LangChain, OpenAI GPT, HuggingFace embedding…

    Jupyter Notebook

  4. Resumai_A-Resume_Analyzer Resumai_A-Resume_Analyzer Public

    Resumai is an intelligent resume analyzer that uses AI to match candidates with job opportunities. This application evaluates resumes against job descriptions using advanced natural language proces…

    JavaScript

  5. RNN-Classification RNN-Classification Public

    This project involves building a sentiment analysis model using Recurrent Neural Networks (RNN) to classify movie reviews from the IMDb dataset as either positive or negative. The IMDb dataset cons…

    Jupyter Notebook

  6. X_Ray_Classifier X_Ray_Classifier Public

    A deep learning-powered web application that automatically detects pneumonia from chest X-ray images using YOLOv8 classification. Built with Streamlit for an intuitive user interface, the system pr…

    Jupyter Notebook