Software Developer
Hey, I'm Aakash 👋
I build reliable backend systems and AI-powered products with Python, Go, FastAPI, Django, PostgreSQL, Redis, and LLM tooling. Always learning and leveling up with every project.
What I Do
Backend Development
Building robust APIs and scalable systems with Python, Go, FastAPI & Django
AI & LLMs
Creating intelligent solutions with RAG, tool calling, and LLM-powered workflows
Systems & Data
Databases, caching, pipelines, and DevOps with PostgreSQL, Redis, Docker
The best way to learn is to build, break things, and build again. Every bug is a lesson in disguise.
About
I'm a backend-focused engineer who enjoys designing scalable systems, clean APIs, and data-driven applications. I've shipped production services with Python and Go, built APIs with FastAPI and Django, and integrated AI workflows into real products.
My day-to-day work covers system design, performance tuning, and building reliable data pipelines. I care about clean architecture, observability, and code that is easy to test, maintain, and scale.
Skills
Technologies and tools I work with to build backend systems and AI-powered products.
Backend
Databases
DevOps
AI
Projects
Reetro Application
Team retrospectives app to capture feedback, vote on themes, and track action items in real time using Golang, PostgreSQL and Redis. Dockerized and ready for deployment.
View on GitHub →AI Agent: Patient Appointment Booking System
This project is a Python-based (FastAPI) patient appointment system that combines standard CRUD flows for doctors, patients, and appointments with an AI-driven Q&A assistant to help users understand and manage scheduling. It can be described as a lightweight AI agent (tool-calling) app.
View on GitHub →Policy AI Agent
A Python FastAPI application that combines organizations, users, leave requests, policies, and an AI chat assistant powered by Cohere with tool/function calling. The AI answers questions using real data from the database and from RAG (Retrieval-Augmented Generation) over policy documents.
View on GitHub →NutriLens
Nutrilens AI is an API that takes food packaging ingredient information (text or image), runs it through AI models, and returns a human-readable explanation of what is inside along with an opinionated health assessment
View on GitHub →Book Recommendation
This is a book recommendation REST API built with Flask that serves two types of recommendations: a popularity-based list (top-rated books) and collaborative filtering (similar books based on user rating patterns). The ML models are pre-trained in a Jupyter notebook using a Books/Ratings/Users dataset and serialized as pickle files, which the API loads at runtime. The app is fully containerized with Docker and exposes endpoints at port 5000.
View on GitHub →Research AI Agent
An autonomous research assistant built with LangGraph, Cohere, FAISS, and Tavily. It searches the live web, retrieves relevant documents from a persistent vector store, and synthesizes a grounded answer using an LLM — all in a single API call. The agent is served via FastAPI and follows an Agentic RAG architecture where each step (search, retrieve, generate) is a node in a stateful graph.
View on GitHub →Standup MCP Server
standup-mcp is an open-source MCP (Model Context Protocol) server that integrates with Claude Code to log daily standups to Notion using plain English. Built with Python, FastMCP, and Cohere's NLP API, it parses natural language messages like 'today I worked on jwt API and dashboard' and automatically appends them as dated bullet lists to a Notion page. It ships with an interactive CLI wizard that handles the full setup — collecting credentials, storing them securely in ~/.zshrc, and generating a credential-free .mcp.json config that safely coexists with other MCP servers. Install it in one command via uv or pipx directly from GitHub, no PyPI publishing required.
View on GitHub →