Arvinth Athi Kesav

AI Engineer specializing in building and deploying agentic AI solutions and RAG pipelines within production environments. Delivered working software using LangGraph, CrewAI, LangChain, FastAPI, and Docker — with strong Python skills, LLM orchestration experience, and a track record of shipping production-ready code end-to-end.

AI/ML Intern at Technical One (2026-01 – 2026-02)

Built real-time AI systems across NLP, Computer Vision, and Explainable AI domains.

• Engineered a Fake News Detection system using DistilBERT fine-tuning, achieving 92% F1-score on benchmark evaluation across Accuracy, Precision, and Recall metrics.
• Deployed a Real-Time Emotion Detection system using CNNs and Transfer Learning (MobileNet/ResNet) for facial expression classification from live video streams at sub-200ms inference latency.
• Architected an Autonomous Fraud Detection system with continual learning and causal explainability (SHAP/LIME) for adaptive real-time risk scoring.
• Implemented Multi-Camera Crowd Behavior Analysis using 3D CNN and CNN-LSTM for spatio-temporal anomaly detection across multi-feed video streams.

B.Tech in Computer Science and Business Systems – MKCE (2022 – 2025)

Advanced Programming Professional & Master Data Science – IIT-M HCL-GUVI (2025-08 – 2026-02)