I'm a passionate full-stack developer and AI enthusiast from NIT Raipur. I love building production-grade systems from scratch, whether it's collaborative drawing apps, intelligent agents, or hybrid recommender systems. I blend deep learning, real-time systems, and solid software architecture to bring ideas to life.

A full-stack Excalidraw-inspired app built from scratch with:

• Next.js, Tailwind, Prisma, PostgreSQL, WebSocket, JWT, Turborepo
• Realtime drawing + chat with authenticated rooms
• CI/CD with GitHub Actions, Docker

End-to-end deep learning + computer vision project:

• Trained NVIDIA-inspired CNN for steering angle prediction
• Used YOLOv8 for lane detection and object tracking
• OpenCV-based simulator to visualize real-time autonomous driving

Spotify-style recommender with:

• Collaborative filtering using cosine similarity on Million Song Dataset
• Content-based filtering via genre, tempo, and pitch features
• Streamlit frontend deployed on AWS S3

• Built with FastAPI, LangGraph, LangChain, and Groq
• LLM-driven pipeline that dynamically routes content generation per language (EN/HI/FR)
• Used LangSmith & LangGraph Studio for debugging

C++, Python, Java, SQL, JavaScript/TypeScript, R

React, Next.js, Redux, Express.js, MongoDB, PostgreSQL, Prisma, WebSockets, JWT, TailwindCSS

TensorFlow, PyTorch, YOLO, LangChain, LangGraph, HuggingFace, OpenAI APIs, Groq, MLflow

Docker, Docker Compose, GitHub Actions, AWS, Streamlit, Flask, FastAPI, Turborepo

Git, Linux, Google Colab, Jupyter, VSCode, Postman, MySQL Workbench

• Email: adime1300@gmail.com
• LinkedIn: linkedin.com/in/aditya-meshram-899955253
• GitHub: github.com/Aditya2600

"I build systems that learn, adapt, and evolve — just like I do."