class AdityaSahani:
def __init__(self):
self.role = "Full-Stack Developer & AI Enthusiast"
self.education = "B.Tech CSE @ IIT Jodhpur - Completed"
self.focus = ["MERN Stack", "Agentic AI", "LLMs", "Deep Learning"]
self.currently = "Building intelligent, full-stack applications"
self.fun_fact = "Tech enthusiast, avid learner, aspiring innovator"
def say_hi(self):
print("Thanks for visiting — let's build something great!")- 🌱 Currently deepening my skills in Full-Stack Development, Agentic AI, and LLMs
- 🤝 Open to collaboration on full-stack and applied-ML projects
- 💬 Ask me about React, Node.js, Python, or AI
- 📫 Reach me at adityasahani893@gmail.com
📍 Japan | 🗓️ Jul 2025 – Apr 2026
- 🇯🇵 Delivered enterprise AI solutions for Japanese clients by developing LLM-powered applications for knowledge management, document intelligence, and business process automation.
- 🏗️ Built an AI agent for automated 2D CAD-to-3D model conversion, combining LLM reasoning with geometric processing to interpret engineering drawings, generate parametric 3D CAD models, and incorporate human-in-the-loop validation for iterative refinement and production-grade accuracy.
- 🧠 Developed Retrieval-Augmented Generation (RAG) systems using LangChain, OpenAI APIs, Google Gemini, and ChromaDB, enabling semantic search and intelligent querying across enterprise knowledge bases.
- ☁️ Designed and deployed scalable AI applications using FastAPI, React, Node.js, Docker, and AWS, integrating vector databases, secure APIs, and cloud-native infrastructure.
📍 India | 🗓️ May 2025 – Jul 2025
- 🤖 Developed agentic workflow systems using FastAPI, Next.js, and LangGraph for real-time AI task orchestration.
- 🖼️ Built AI-powered image generation and auditing pipelines for e-commerce, including masking, defect detection, and automatic regeneration.
- 📈 Improved generated image quality by 35% through automated validation and redetailing workflows.
📍 Asia (Remote)
- 🌱 Built and maintained full-stack web applications using Django, React.js, SQL, and AWS.
- 📊 Developed scalable backend APIs, database models, and responsive dashboards for environmental sustainability solutions.
- ☁️ Improved application performance and deployment workflows while collaborating with cross-functional teams.
My Some of the Best Projects
| Project | Description | Tech |
|---|---|---|
| PromptChef - VSCode Extension | Turn vague prompts into clear, well-specified ones — right inside VS Code. | TypeScript · VSCode · LLMs |
| Walmart-AI : Shopping Assistant | AI-powered shopping assistant that revolutionizes the online shopping experience for customers by automating online product searching - giving users with best ranked products based on his demand and saves time. It combines advanced AI capabilities, deep product research, voice interaction, and seamless checkout processes to create an unparalleled e-commerce platform. | NextJS · TypeScript · Gemini · SerpAPI |
| MusicRecommendationBasedOnFacialExpression | A web application that recommends music based on the facial emotion of the user - right into the website from Spotify. It utilizes ReactJS for the frontend, Node.js and Express for the backend, and MongoDB for the database. Additionally, it integrates PyTorch and TensorFlow for facial emotion recognition. | ReactJS · NodeJS · MongoDB · PyTorch · TailwindCSS |
⭐️ From Aditya Sahani — thanks for stopping by!