Building intelligent
AI solutions.

I am Amith Ajith, a Backend Software Engineer focused on Generative AI and Multi-Agent Systems.

View Work

About Me

I am a Backend Software Engineer with expertise in Python, Generative AI and Azure. Currently working at LTIMindtree in Rotterdam, I focus on building advanced RAG-based chatbots and multi-agent/multi-modal systems using Autogen, Microsoft Agent Framework and Google ADK.

My background includes extensive work with Azure, GCP, and backend frameworks like FastAPI and Flask, while continuously deepening my expertise in Data Structures & Algorithms and SQL.

Want to know more? Try the AI assistant in the bottom right corner to explore my work and navigate the site.

Languages

PythonSQLJavaScriptC++CJava

Frameworks & Libraries

AutogenFastAPIFlaskReactJSPandasSQLAlchemyStreamlit

Cloud & Tools

Azure OpenAIGCP Vertex AIGitDockerJiraDevOps

Experience

Specialist - Data Sciences

Jan 2025 - Present
LTIMindtree (Rotterdam, Netherlands)

Deploying customer-facing RAG-based chatbots to global markets and prototyping real-time voice assistants and agentic chatbots. Transformed legacy chatbots into scalable agentic systems using Autogen, improving latency by 50%. Contributed to Microsoft Autogen repository. Built advanced NL to SQL solutions.

Data Scientist

July 2021 - Dec 2024
LTIMindtree (Bengaluru, India)

Worked with a leading FMCG client's innovations team on AI initiatives. Designed NL2SQL-based GenAI chatbot using GPT-3.5. Developed REST APIs in Flask and automated report generation. Member of MLOps CoE, developed pipelines in Vertex AI. Built model explainability tools using ReactJS.

Selected Works

Building This Portfolio & Ami

Building This Portfolio & Ami

Full Stack & AI

A deep dive into creating an agentic portfolio website with a sarcastic AI assistant using Next.js, FastAPI, and Claude.

Voice Agents

Voice Agents

AI & Telephony

Developed a tool for creating voice agents that place automated phone calls using Twilio and gpt-realtime-based voice assistant.

AutoGen Multi-Agent Streaming

AutoGen Multi-Agent Streaming

AI & Multi-Agent Systems

Contributed a sample to Microsoft's AutoGen demonstrating streaming chat with multi-agent handoffs using FastAPI and autogen-core.

Tool Failure Prediction

Tool Failure Prediction

Machine Learning

ML models to predict tool failure in Friction Stir Welding processes using thermodynamic model datasets.

Let's work together.

I'm currently available for freelance projects and open to full-time opportunities.

Get in Touch