I am an AI & Machine Learning student at VIT Bhopal University focused on building practical, reliable AI systems. My interests span Machine Learning, Deep Learning, Computer Vision,NLP and software engineering. I enjoy turning ML research ideas into usable products through clean APIs, modular architectures, model optimization, and thoughtful product engineering.
Open To: AI/ML Engineering • Software Engineering • Computer Vision • Applied AI • Research & Development
B.Tech — Artificial Intelligence & Machine Learning
VIT Bhopal University
Libraries & Systems: Scikit-learn • Hugging Face • ChromaDB • Neo4j • Grad-CAM • CLAHE • Optical Flow
Cloud / Edge: Cloudflare Workers • NVIDIA Jetson • CUDA • TensorRT
| Domain | Proficiency | Details |
|---|---|---|
| Machine Learning | Strong | Model development, evaluation, transfer learning and applied ML |
| Deep Learning | Strong | PyTorch, TensorFlow, CNNs, transformers and model optimization |
| Computer Vision | Strong | OpenCV, image preprocessing, Grad-CAM, optical flow and real-time CV |
| Medical AI | Strong | Chest X-ray analysis, DICOM workflows and explainable predictions |
| NLP | Intermediate | Transformer-based text classification and emotion analysis |
| Edge AI | Intermediate | NVIDIA Jetson, CUDA, TensorRT and low-latency inference |
| RAG | Learning | Embeddings, vector databases, chunking, retrieval and RAG pipelines |
| Generative AI | Exploring | Exploring LLM-based application development |
| Agentic AI | Exploring | Exploring agent workflows and tool-using AI systems |
🫁 PulmoScan — AI-Powered Chest X-ray Pneumonia Detection
An explainable medical-AI pipeline for pneumonia detection from chest X-rays. The system uses transfer learning with ResNet-50, image preprocessing, uncertainty estimation and Grad-CAM explanations. A FastAPI backend exposes REST endpoints for DICOM/image processing and automated PDF report generation, while Docker provides reproducible deployment.
| Area | Details |
|---|---|
| Stack | Python • ResNet-50 • PyTorch • OpenCV • FastAPI • Docker |
| Scale | Image classification and DICOM-compatible inference workflow |
| Performance | 90–94% test accuracy; ROC-AUC 94.2% |
| Explainability | Grad-CAM visualizations |
| Reliability | Uncertainty estimation, automated testing and modular CLI tools |
| Impact | Converts a deep-learning model into an explainable API-driven application |
| Repository | GitHub |
📊 Emotion Drift Detection — Enterprise Sentiment Analytics
An NLP analytics system designed to track changing customer emotions over time. A fine-tuned DistilBERT model processes text while the backend provides secure REST APIs for real-time sentiment analytics. The application supports emotion trajectories, SLA-risk analysis and deployment through cloud and desktop environments.
| Area | Details |
|---|---|
| Stack | Python • DistilBERT • FastAPI • React • Electron |
| NLP | Fine-tuned transformer-based emotion classification |
| Analytics | Emotion trajectory and sentiment-drift analysis |
| Backend | Secure REST API architecture |
| Deployment | FastAPI • Hugging Face Spaces • Docker |
| Impact | Converts individual sentiment predictions into time-aware customer insights |
| Repository | GitHub |
🧠 Enterprise Knowledge Copilot — Nasscom Hackathon 2026
A graph-enhanced enterprise RAG system developed with Team Runtime Terror. The architecture combines a vector store with a graph database to support semantic retrieval and relational reasoning. LangGraph coordinates the agent workflow while LangChain connects retrieval and tools.
| Area | Details |
|---|---|
| Stack | LangGraph • Gemini 1.5 Pro • ChromaDB • Neo4j • LangChain |
| Retrieval | Vector-based document retrieval with ChromaDB |
| Reasoning | Neo4j-based entity and relationship queries |
| Architecture | Document ingestion → embeddings → retrieval → grounded response generation |
| Contribution | End-to-end RAG flow and relational reasoning module |
| Event | Nasscom Hackathon 2026 |
| Repository | GitHub |
🌱 AgriBot — AI-Powered Agricultural Chatbot
An AI-powered agricultural chatbot designed to recommend suitable crops based on soil, climate, and environmental conditions. The system uses machine learning to provide data-driven insights for better farming decisions.
| Area | Details |
|---|---|
| Stack | Python • Scikit-learn • HTML • CSS • JavaScript |
| Machine Learning | Support Vector Machine (SVM) |
| Preprocessing | MinMaxScaler • StandardScaler |
| Prediction | Crop recommendation based on soil & climate conditions |
| Interface | Interactive chatbot-based web interface |
| Impact | Data-driven agricultural guidance for crop selection |
| Repository | GitHub |
💡 Visual Stress Detection System
A computer-vision and biomedical signal-processing system that analyzes facial micro-movements using dense optical flow. The project explores relationships between visual micro-tremors and physiological stress indicators.
| Area | Details |
|---|---|
| Stack | Python • OpenCV • Optical Flow • Signal Processing • Scikit-learn |
| Input | Facial motion captured at 60Hz |
| Method | Dense optical-flow analysis and signal processing |
| Result | 86% correlation with standard HRV indicators |
| Impact | Explores non-contact visual approaches to physiological signal analysis |
May 2026 – June 2026 · Remote
Contributing to an AI-powered photo culling platform, working across backend services and desktop application integration.
- Developed and integrated backend services using Cloudflare Workers and scalable REST API design.
- Worked across React, TypeScript, Electron and Node.js integration layers.
- Contributed to AI-driven photo-culling functionality and model-driven automated photo selection.
- Debugged integration issues and improved API reliability and performance.
- Collaborated through feature design, sprint planning, code reviews and testing.
Skills: Cloudflare Workers • REST APIs • TypeScript • React • Electron • Node.js • AI Product Engineering
| Recognition | Details |
|---|---|
| 🏆 Nasscom Hackathon 2026 | Hackathon participant — Team Runtime Terror |
| 💻 LeetCode | Continuous DSA and problem-solving practice |
| 🤖 Applied AI Projects | Built projects across Medical AI, NLP, Computer Vision and Edge AI |
- Introduction to Machine Learning
Certificate
- Fundamentals of Computer Vision
Certificate
- Cloud Computing
Certificate
- Java — Skill Certification
Certificate - SQL — Skill Certification
Certificate
- Applied Machine Learning in Python
Certificate
- ServiceNow Virtual Internship Program
Certificate
- GenAI Powered Data Analytics Job Simulation
Certificate
Learning:
- RAG Systems
- Vector Databases
- Embeddings & Semantic Search
- Document Chunking & Retrieval
- Generative AI
Building:
- Applied AI/ML systems
- Computer Vision pipelines
- API-driven ML applications
Exploring:
- Agentic AI
- Advanced RAG architectures
Open To:
- AI/ML Engineering
- Software Engineering
- Computer Vision
- Applied AI
- Research & Development
“Build systems that are intelligent, reliable, and useful.”