AI Builder of the Week

I'm a Gen AI Developer based in India. I build applied AI systems — think LLM-powered agents, production APIs, and agentic workflows. On the engineering side, I work with Python, FastAPI, LangChain, and AgentScope. I also do React Native and web frontend when needed. Outside of coding, I write technical content breaking down GenAI concepts for other developers. I believe in learning by building and sharing what I learn along the way.

I was always into programming, but GenAI grabbed me when I first experimented with LLMs and saw how they could be composed into real applications — not just chatbots, but actual systems that reason, retrieve, and act. I started building small projects, then joined Kactii Academy's bootcamp which pushed me to go deeper. Building RAG pipelines, working with vector databases, and deploying agents in production is what I do day-to-day now.

I'm working on a few things. One is an agentic AI system using LangChain and AgentScope that can orchestrate multiple tools to complete complex tasks. I'm also contributing to Kactii Academy's platform itself — building features with FastAPI and MongoDB. And I maintain a technical blog where I write about GenAI patterns, LLM tooling, and practical agent development. I try to push code every day.

Python is my main language. FastAPI for APIs — it's fast, clean, and async-native. LangChain for LLM orchestration and agent building. MongoDB with vector search for data storage. N8N for workflow automation when I need low-code pipelines. For deployment, I use Vercel for web apps and Ubuntu servers for heavier workloads. I've also been exploring datasets and fine-tuning lately.

I believe the best way to learn is to explain. I write technical articles and tutorials that break down GenAI concepts — things like how to build RAG systems, how agents work under the hood, prompt engineering patterns, and production deployment tips. I publish on my personal site and share notes for the community. Writing forces you to understand things deeply, and it helps other developers who are on the same path.

Start building immediately. Don't just watch tutorials — pick a problem, build an end-to-end solution, and deploy it. Even if it's rough, you'll learn 10x more than following along passively. Also, get comfortable with the fundamentals: understand how embeddings work, how retrieval works, how LLMs reason. And contribute to a community — Kactii has been incredible for me because you're surrounded by people building real things, not just talking about AI.

I want to go deeper into multi-agent systems — building agents that can collaborate, delegate, and self-correct. I'm also interested in pushing the boundaries of what you can do with open-source LLMs in production. Long term, I want to build AI products that solve real problems and keep writing content that helps the developer community level up.