AI Engineer & Full Stack Developer
I transform complex data into intelligent solutions.
Specializing in ML/DL systems, federated learning, and production-ready web applications that drive real business impact.
I'm an AI Engineer and Full Stack Developer passionate about leveraging artificial intelligence and machine learning to create innovative solutions. My expertise spans across Machine Learning, Deep Learning, Data Science, Agentic AI, and modern web development.
With hands-on experience in Python, React, Next.js, and Flutter, I build intelligent applications that bridge the gap between cutting-edge AI technology and user-friendly interfaces.
Years Experience
Projects Completed
GitHub Repos
Supervised & unsupervised learning, model optimization
Neural networks, CNNs, RNNs, and transformers
Data analysis, visualization, and statistical modeling
Autonomous agents and intelligent automation
TensorFlow, PyTorch, Scikit-learn, Pandas, NumPy
Modern web apps with SSR and optimal performance
HTML, CSS, JavaScript, responsive design
Cross-platform mobile and web applications
A commercial web application that enables users to download entire websites for offline use. Professional tool for archiving and offline browsing.
Distributed ML platform with Docker, WebSockets, and federated averaging. Features multiple local servers contributing to a shared global model while maintaining data privacy.
Fine-tuned Mistral 7B LLM using LoRA for insurance FAQ domain. Advanced parameter-efficient fine-tuning of large language models.
Intelligent document explainer using LLM APIs and vector databases. Built with React and Vite for fast, interactive learning assistance.
Real-time one-to-one chat application with Node.js, Express, Socket.IO, and PostgreSQL. Features user authentication, chat history, and live messaging.
Full-featured digital marketplace platform for buying and selling digital products. Modern e-commerce solution with responsive design.
Streamlit application for building AI models without coding. Features data processing, model fitting, and testing - making ML accessible to everyone.
Collection of ML projects including fraud detection, taxi fare prediction, genetic algorithms, and optimization systems. Explore more on GitHub.
Feedback from collaborators and clients
"Arpan's work on the federated learning system was exceptional. His deep understanding of distributed ML and clean code architecture helped us build a scalable solution."
"The AI NoCode platform Arpan built makes machine learning accessible to our entire team. Impressive full-stack skills combined with ML expertise."
"Professional, reliable, and innovative. Arpan delivered Website2Zip with excellent attention to detail and user experience."
Have a project in mind? Let's work together!
arpan.g2102@gmail.com
@ArpanGoswami515
Kolkata, India