Harinand P

About

Computer Science undergraduate specializing in Artificial Intelligence, Machine Learning, and Computer Vision, with hands-on experience building and deploying real-time computer vision, edge AI, and embedded systems using Python and modern ML frameworks. Also experienced in UAV and autonomous aerial systems, including aircraft design, avionics, and flight autonomy.

Experience

• Android App Development Intern
  CL Infotech Pvt. Ltd.
  Feb 2026 – May 2026

Education

• B.E. in Computer Science and Engineering
  Jyothy Institute of Technology
  Computer Science and Engineering · Dec 2022 – Jul 2026

Skills

• SQL
• C
• Python
• Pandas
• Streamlit
• scikit-learn
• OpenCV
• PyTorch
• TensorFlow
• ESP32
• Flutter
• Keras
• LangChain
• Yolo
• Jetpack Compose
• RAG
• LLMs
• ONNX
• Mission Planner
• Shap
• ArduPilot

Projects

• DocIQ — Local RAG Document Intelligence ↗
  Python, LangChain, ChromaDB, Ollama, Streamlit

  Developed a fully local retrieval-augmented generation (RAG) system using LangChain, ChromaDB, and Ollama (Llama 3.2), enabling natural language Q&A over any PDF with cited answers. Integrated MMR retrieval, few-shot prompting, and chain-of-thought reasoning to improve factual accuracy of generated answers.

• Edge-Powered XAI Crop Recommendation System ↗
  ONNX, ESP32, Flutter, SHAP, LIME

  Developed an edge-powered crop recommendation system using IoT soil sensors and ESP32, with an ONNX-based ML model embedded directly in a multilingual Flutter app for fully offline inference. Integrated explainable AI (SHAP & LIME) for interpretable crop insights and real-time dashboards; recognized with Best Research Paper at NCRIE-2025, K.S. Institute of Technology.

• EO/IR Video Classification ↗
  Python, YOLOv8, Gradio, Raspberry Pi

  Built a real-time military entity detection system using YOLOv8 and YOLOv8-Nano on dual-spectrum EO/IR video feeds, classifying 7 military entity types with 85–88% accuracy. Automated Critical/High/Medium/Low threat assessment via a Gradio-based side-by-side surveillance UI; optimized for edge deployment on Raspberry Pi, achieving 28–30 FPS inference throughput.

Courses & certifications

• Machine Learning Specialization (Supervised Machine Learning, Unsupervised Machine Learning, Advanced Learning Algorithms) ↗ · DeepLearning.AI & Stanford University · 2025

🏆 Achievements & awards

• Best User Interface Award · 2024

  Enigma 2024, JIT (Stock Market Predictor project).

• Best Research Paper · 2025

  "Edge-Powered XAI Based Crop Recommendation System," National Conference on Recent Innovations in Engineering (NCRIE-2025), K.S. Institute of Technology, Bengaluru, Dec 2025.