I have recently delivered four notable AI‑focused projects that blend large‑language models, tool orchestration, and domain‑specific data pipelines.
Cost‑Aware Multi‑Model AI Router with MCP Agent Orchestration (Oct 2026 – Sep 2026): Built an intelligent AI routing platform that dynamically selects Gemini or Mistral based on task complexity, confidence, cost, and fallback requirements.
Implemented LangChain, LangGraph, MCP, Clerk OAuth, and LangSmith for agent orchestration, observability, and secure user‑scoped tool access.
Integrated dynamic MCP connectors for GitHub, Gmail, and Google Drive, enabling multi‑step tool calling and cross‑connector workflows while tracking cost, savings, confidence, and model performance.
Telco Customer Churn Prediction & Segmentation (Oct 2026 – Sep 2026): Built an end‑to‑end churn prediction pipeline using Python, Scikit‑learn, and XGBoost on 7,043 customer records.
Applied hyperparameter tuning and SHAP to improve model performance and explain key factors influencing churn.
Used K‑Means clustering to segment customers and developed a churn‑risk and business‑priority framework.
GenZ Voice – Voice AI Healthcare Assistant (Sep 2026 – Jun 2026): Developed a real‑time Voice AI healthcare assistant as part of Team GenZ Voice for Voice‑A‑Thon 2026, organized by SnapServe AI × Vobiz AI.
Integrated a voice agent with backend APIs to handle AI conversations, location‑based healthcare assistance, report generation, and email delivery.
Secured 1st Place in Round 1 among participating teams.
SportsGPT – AI‑Powered Sports Assistant: Built an AI‑powered sports assistant that combines Large Language Models (LLMs), Retrieval‑Augmented Generation (RAG), and Model Context Protocol (MCP) to deliver intelligent and up‑to‑date sports information.
Developed a FastAPI‑based backend with MongoDB vector search for knowledge retrieval and MCP tools integrated with Tavily for current sports news and updates.
Implemented intelligent tool calling, query routing, multi‑step tool execution, and LLM fallback mechanisms using Groq, Gemini, and LiteLLM.
Hello everyone! 👋 I’m a Mechatronics Engineering graduate passionate about Software Engineering, Generative AI, Machine Learning, and intelligent automation. I enjoy exploring new technologies and building AI‑powered applications that solve real‑world problems.
Technical Skills: Programming: Python, SQL, JavaScript, TypeScript
Career Interests: I'm interested in opportunities involving Generative AI, Applied AI, AI/ML Engineering, Python Backend Development, LLM Application Development, and AI Workflow Automation.