I'm an AI/ML + Backend Engineer based in India, mastering Python and building intelligent systems — agents, pipelines, and APIs that actually do something useful.

My philosophy is simple: automate the boring, ship the interesting, and write code that makes me look like I worked harder than I did.

I build AI-powered systems that turn complex workflows into automated experiences.

Recently, I've been focused on agentic AI, designing applications that can research, reason, and execute tasks with minimal human intervention.

Beyond machine learning models, I enjoy building the backend systems that make AI applications reliable in production—from APIs and orchestration layers to databases, asynchronous workflows, and deployment infrastructure.

Right now, I'm exploring multi-agent architectures, tool-calling systems, memory-enabled agents, and autonomous workflows, with the goal of building AI products that do more than answer questions—they get things done.

AI/ML — Python · TensorFlow · Keras · scikit-learn · ONNX · Hugging Face · MLflow · LangChain · LangGraph · CrewAI

Backend — FastAPI · Django · Express · Node.js · REST APIs · async Python

Data — PostgreSQL · MongoDB · MySQL · Redis · Firebase · Supabase

Infra & Tooling — Docker · Git · Postman · Jupyter

Frontend (when needed) — React · Next.js · Tailwind CSS

• An Agentic Travel Planner - A Project that turns a planning desire prompt to a researched planned itenary within seconds.
• An AI image detector - A project that takes images, batch infers them and send a report straight to your email with details about which images are ai influenced.

GitHub · LinkedIn · Kaggle · LeetCode · debankurdutta04@gmail.com · Portfolio

Ask me about ML, overfitting, or the hustle it takes to get an offer.