I'm Karthik Nadar, a passionate Backend, AI, and Cloud Engineer. I thrive at the intersection of scalable infrastructure and cutting-edge machine learning, and I’m always exploring ways to bridge traditional systems with intelligent, agentic solutions.

I started my journey in full-stack and backend development — building resilient architectures using Express.js, FastAPI, and Python, and crafting frontend interfaces with React and Next.js. Over time, I developed a strong foundation in DevOps and Cloud Infrastructure, which helped me design systems that scale and stay reliable in production.

Now, my focus is on building intelligent systems — combining the best of AI, automation, and backend architecture.

My cloud and infrastructure toolkit includes:

I enjoy building robust CI/CD pipelines, automating workflows, and using orchestration tools to streamline development and deployment.

• Deep Learning: Built and fine-tuned CNN models for computer vision tasks
• Generative AI: Worked with GANs, VAEs, energy-based models, and diffusion models
• Text-to-Image: Fine-tuned diffusion models using LoRA, DreamBooth, and custom pipelines for creative generation

I'm deeply involved in building AI agents and agentic systems — crafting workflows that go beyond simple prompt-response setups. I’ve worked extensively with:

• LangGraph, LangChain, LangSmith
• Vercel AI SDK for full-stack agent deployment
• Vector databases like Pinecone for fast and accurate RAG pipelines
• Local LLM orchestration using Ollama
• Designing and optimizing chunking strategies and retrieval flows

This is where I love to geek out — building smart, modular systems that reason, plan, and act. I see agents as the next layer of app intelligence, and I’m all-in on exploring that space.

A platform that lets you converse with your database in natural language and instantly turn complex data into clear, actionable insights. It’s all about making data more approachable, conversational, and useful.

Visit Plum AI An AI-powered search engine (inspired by Perplexity) using Tavily Search API and OpenAI, offering fast, contextual, and accurate responses.

Visit KitKat AI A smart note-taking app that lets you chat with your notes. Powered by conversational AI and natural language search to help you organize and retrieve ideas effortlessly.

Visit Avocado AI A text-to-video app that uses Google TTS, Firebase Storage, and Remotion.js to turn written content into dynamic videos. Built with Clerk, PostgreSQL (Neon), and Drizzle ORM.

Check out my GitHub for more on:

• Backend Engineering with Python, Express.js, FastAPI
• Frontend Development with React, Next.js
• DevOps and Cloud Infrastructure with AWS, Docker, Kubernetes
• AI Systems with LangChain, LangGraph, Pinecone, Ollama
• Text-to-Image AI and generative models

Let’s connect on LinkedIn or feel free to explore my GitHub for more of my work.

Thanks for stopping by — always down to connect, collaborate, or build something cool together 🚀