Mitul Khemani

Machine Learning & AI Developer

Building intelligent, scalable, and data-driven systems that solve real-world problems. Passionate about transformers, retrieval-augmented generation, and production ML.

About Me

I'm a pre-final year engineering student at The LNM Institute of Information Technology, Jaipur, with a strong focus on machine learning, deep learning, and AI systems. My journey has been driven by a passion for applying cutting-edge ML techniques to solve real-world problems.

Throughout my studies, I've built end-to-end systems spanning NLP, computer vision, time series analysis, and generative AI. What excites me most is the intersection of data science and software engineering—taking research ideas from paper to production.

I'm a continuous learner with strong fundamentals in mathematics and algorithms. I thrive in environments where I can tackle complex problems, collaborate with talented teams, and contribute to building intelligent systems that make an impact.

🎓

Education

B.Tech in Electronics & Communication

The LNM Institute of Information Technology, Jaipur

Aug 2023 - May 2027 | CGPA: 8.16

⭐

Recognition

Branch Topper - Consistent Academic Excellence

Hackathon Finalist - NIT Delhi (Guru Setu Project)

Skills & Expertise

Machine Learning & AI

BERTGPTTransformersNLPRAG ModelsDeep LearningComputer Vision

Data Science & Analytics

Time Series AnalysisXGBoostFeature EngineeringStatistical ModelingData PipelinesETL

Programming & Development

PythonC++SQLFastAPIFlaskJavaScript

Tools & Frameworks

PyTorchTensorFlowScikit-learnPandasGitDockerAWSLinux

Featured Projects

Auto Healing Customer Chat Support System

End-to-end customer support automation using transformer models for intelligent issue resolution

The Problem

Manual customer support is time-consuming and inconsistent. Businesses need an intelligent system that can resolve common issues automatically while seamlessly escalating to human agents when needed.

The Solution

Built an AI-powered chat system using BERT for intent classification and GPT for contextual response generation, with automated ETL pipelines for continuous learning from past interactions.

Impact & Results

~80% classification accuracy, automated issue resolution, intelligent escalation

BERTGPTFastAPIETL PipelinesPython

View on GitHub

E-Commerce Competitor Price & Stock Recommendation

Automated pricing intelligence system with daily market analysis and demand prediction

The Problem

E-commerce businesses struggle to stay competitive with dynamic pricing. Manual price monitoring is inefficient and lacks predictive insights.

The Solution

Developed an automated system that scrapes competitor data daily, feeds it to an XGBoost model for demand prediction, and provides actionable pricing recommendations.

Impact & Results

82% accuracy, 60% reduction in deployment time via Docker CI/CD, real-time recommendations

XGBoostWeb ScrapingDockerCI/CDStreamlit

View on GitHub

Real-Time Stock Price Prediction System

Time-series forecasting for both long-term and short-term stock price predictions

The Problem

Investors need reliable, real-time stock predictions with both daily and intraday forecasts for informed trading decisions.

The Solution

Created an LSTM-based system with separate workflows for next-day and next-minute predictions, featuring dynamic retraining and automated data ingestion.

Impact & Results

Real-time predictions across any stock ticker, automated daily retraining, live deployment on Render

LSTMTime SeriesFlaskYFinancePython

View on GitHub

Guru Setu: AI Teacher Support System

Comprehensive AI assistant for educators with RAG, voice integration, and curriculum support

The Problem

Teachers need intelligent assistance with content generation, planning, and interactive teaching tools. Manual preparation is time-intensive.

The Solution

Developed an integrated platform combining Generative AI with RAG models, voice-based interaction via Whisper, and Google OAuth authentication.

Impact & Results

80% response relevance on 50+ real-world queries, voice-enabled interface, hackathon finalist recognition

Generative AIRAG ModelsWhisperGoogle GeminiGoogle OAuth

View on GitHub

Achievements & Strengths

Team Leader and Hackathon Finalist at NIT Delhi for Guru Setu - selected among top teams for technical innovation

Branch Topper in Electronics and Communication Engineering - consistent academic excellence

Strong ML fundamentals with practical end-to-end project execution across multiple domains

Demonstrated ability to bridge data science and software engineering in production environments

Resume & Details

Download my full resume to see a comprehensive overview of my experience, projects, and accomplishments.

View Resume

Let's Connect

I'm actively looking for internships and entry-level opportunities in AI/ML and Data Science. Whether you have a project in mind, a question, or just want to chat about AI and tech, I'd love to hear from you.

Email Me LinkedIn GitHub

📧 khemanimitul@gmail.com | 📱 +91 7470924433