KATINEEDI KARTHIK

About

B.Tech student in Artificial Intelligence and Machine Learning with project experience in RAG-based document analysis, predictive modeling, and API deployment. Currently interning in AI, with hands-on work in computer vision, model benchmarking, and full-stack ML integration.

Experience

• Artificial Intelligence Intern
  GGF Astraa Analytica · Hyderabad
  Aug 2026 – Present

  Optimized YOLO object detection models for real-time identification and localization in autonomous navigation systems, improving inference speed by 60%. Evaluated and benchmarked model performance across 30 test scenarios, identifying edge cases and contributing to pipeline integration alongside the AI team.

Education

• B.Tech
  CMR Engineering College
  AI/ML · 2023 – 2027
• Intermediate
  Sri Chaitanya Junior College
  MPC · 2021 – 2023
• SSC
  SRM High School
  2021

Skills

Tools / apps / platforms

Projects

• Customer Churn Prediction
  Python, XGBoost, FastAPI

  Developed a customer churn prediction model using Python and Scikit-learn/XGBoost, achieving 90% accuracy by analyzing customer behavior and transaction data. Performed feature engineering and EDA to identify key churn drivers, improving model performance and enabling data-driven retention strategies. Deployed the model as a REST API using FastAPI with database integration (PostgreSQL) for real-time predictions.

• Tender analysis system
  Python, LangGraph, OCR, RAG, FastAPI

  Built a multi-agent tender analysis system that extracts text from scanned/PDF tender documents via OCR and uses a RAG pipeline to evaluate technical and commercial requirements against company capabilities. Automated generation of techno-commercial compliance reports covering technical compliance, commercial terms, key requirements, and gap analysis, cutting manual review time by 80% and reducing report turnaround from 2 hours to 20 minutes.

• AI Medical Report Analyzer
  Python, LangChain, PyMuPDF, RAG, FastAPI

  Built an AI-powered medical report analyzer using a RAG pipeline to parse 20+ clinical parameters from PDF reports via PyMuPDF and flag values outside reference ranges with 89% accuracy. Designed a persistent patient history store that retrieves prior reports and combines them with current results to generate personalized dietary recommendations, reducing manual review time by 70%.

Courses & certifications

• Machine Learning course · IBM

🏆 Achievements & awards

• Special prize for innovative solution and technical excellence

  Awarded for a hackathon conducted by Osmania University.

⚽ Extracurricular activities

• Competitive Programming

  Actively participating in contests conducted by LeetCode and CodeChef.

• Hack-AI-Thon participant

  Participated in Hack-AI-Thon, collaborating in a team to design and prototype an AI-based solution under time constraints.