Karthik Kadam - AI Software Engineer
Karthik Kadam - Tech Lab
Karthik Kadam - Innovator
Karthik Kadam - Original Photo
drag or click to slide

hi karthik here. 👋

22yo software engineer from Davangere, Karnataka 🇮🇳

Backend by profession, AI systems by passion. Specialized in Agentic AI (MCP & RAG), LLM Fine-Tuning, and scalable full-stack web applications.

For Q&A, start a chat with Karthik Support

Resume
  • IngestLab Logo IL

    IngestLab

    Software Engineer (Full-Time)

    • Worked on Langflow and optimized AI workflows using Retrieval-Augmented Generation (RAG) with vector search implementation in Qdrant, improving the response speed of the AI by 65%.
    • Worked on AWS and Azure pipelines improving data collection and transfer by 14%.
    • Fine-tuned DeepSeek-Coder model using Unsloth and LoRA techniques, optimizing training efficiency, reducing GPU memory consumption, and accelerating model experimentation and deployment.

    Software Engineer (Intern)

    • Architected and deployed Model Context Protocol (MCP) servers for integrating and managing Google Analytics 4 and Meta platform data workflows.
    • Achieved product mastery through hands-on development, rigorous experimentation, and rapid feature delivery.

technical skills

Languages
Python Java JavaScript HTML5 CSS3 SQL
AI & LLMs
RAG Model Context Protocol (MCP) DeepSeek-Coder Unsloth & LoRA LangFlow Qdrant Vector Search Google Gemini API Ollama TensorFlow.js
Backend & Web
Node.js Express.js Flask WebSocket REST APIs
Databases & Cloud
MongoDB MySQL Qdrant DB AWS Azure Docker Git

featured projects

github
Natural Language to SQL Query Generator

Natural Language to SQL Query Generator

Fine-tuned DeepSeek-Coder using Unsloth to generate high-accuracy SQL queries from natural language prompts. Quantized model to GGUF for efficient local inference with Hugging Face datasets.

DeepSeek-Coder Unsloth LoRA GGUF Python Hugging Face
Colab Source
Model Context Protocol (MCP) Server

Model Context Protocol (MCP) Server

Architected and deployed high-performance Model Context Protocol (MCP) servers for integrating and managing real-time data workflows across Google Analytics 4 (GA4) and Meta platforms.

MCP Python / Flask Google Analytics 4 Meta API Render
Live API Source
Code Studio AI IDE

Code Studio: AI-Powered Dev Environment

Real-time collaborative pair programming environment powered by WebSockets, Google Gemini & local Ollama models for inline code completions, with multi-language compiler and terminal integration.

Express.js WebSocket TensorFlow.js Gemini API Ollama
Source
Pip Manager

Pip Manager (Desktop GUI Tool)

Intuitive desktop GUI tool for developers to manage python pip installations efficiently, including sorting packages by size, date, and version with one-click dependency upgrades.

Python Tkinter Pkg_resources Desktop GUI
Website Source
Code Quest 1.0 Competition Platform

Code Quest 1.0 Competition Platform

Inter-college technical competition portal engineered for BIET Davangere, handling over 100 participant team registrations, event guidelines, and leaderboard coordination.

Node.js Express HTML5 / CSS3 Event Portal
Website Source

certifications & honors

get in touch

Have an exciting engineering role, freelance project, or collaboration idea? Send a message directly or connect via LinkedIn.

Hi! 👋 I'm Karthik's AI Support assistant. Ask me anything about his work at IngestLab, DeepSeek-Coder fine-tuning, RAG pipelines, education at BIET, or how to reach him!