- 🎓 CSE Undergraduate (2026) passionate about AI/ML
- 🚀 Love taking projects from idea → prototype → production
- 🧠 Focused on clean architecture, scalable backends, and real-world ML systems
- 🤖 Currently working as an AI/ML Intern, building LLM-powered solutions
- 📫 Reach me at: dhruvkjani@gmail.com
- ⚡ Fun fact: I name variables better than I name Wi-Fi networks.
Tech Stack: Python, Flask, TF-IDF, Naive Bayes, Pandas
A production-grade health recommendation engine that analyzes user symptoms and demographics to generate risk-aware health suggestions.
Key Features:
- NLP-based symptom processing using TF-IDF vectorization
- Probabilistic predictions with confidence scoring
- Demographic-aware tailoring and chronic-risk flagging
- Explainable AI design for real-world medical contexts
Tech Stack: PyTorch, Torchaudio, Librosa, Modal, Next.js, Tailwind CSS
End-to-end audio classification system using a ResNet-inspired CNN trained on ESC-50 dataset (50 classes), with an interactive web interface.
Key Features:
- Mel-spectrogram-based feature extraction for audio processing
- Serverless inference using Modal for production scalability
- Real-time audio upload with top-3 predictions and confidence scores
- Interactive visualizations of waveforms, spectrograms, and feature maps
Tech Stack: React, Node.js, Express, JWT, Google Gemini API
Secure, full-stack AI application that generates professional email replies using large language models.
Key Features:
- JWT-based authentication with role-based access control
- Input validation and secure API architecture
- LLM integration for context-aware email responses
- Optimized React components for performance and UX
Tech Stack: Node.js, WebSockets, Tailwind CSS
Secure, real-time chat system with token authentication and encrypted message storage.
Key Features:
- WebSocket-based real-time bidirectional messaging
- Token-authenticated socket connections for security
- End-to-end encrypted message storage
- Responsive, mobile-first UI design
- 🤖 Contributing to applied AI/ML solutions as an AI Intern
- 🧩 Building production-ready applications with LLMs (Large Language Models)
- 🔍 Designing RAG (Retrieval-Augmented Generation) pipelines for context-aware AI systems
- 🗂️ Exploring vector databases & embeddings for semantic search and knowledge retrieval
- 🎯 Improving prompt engineering, evaluation, and optimization for LLM-based workflows
- 🧪 Applying ML concepts to production environments, not just notebooks
|
|
⭐ Star my repositories if you find them interesting!