• 🔭 I'm currently working on Computer Vision & AI Solutions
• 🌱 I'm currently learning Artificial Intelligence, Deep Learning, Computer Vision
• 👯 I'm looking to collaborate on Open Source Projects
• 💬 Ask me about AI, ML, Data Science, Data Engineering, Cloud, Solutions Architect
• 📫 How to reach me: LinkedIn | chandrasekhar090591@gmail.com
• ⚡ Fun fact: I love hiking and going on long drives 🚗⛰️

Description: A production-grade agentic AI built entirely from scratch on the raw Anthropic SDK — no LangChain, no Agent SDK. Implements a manual tool-use loop where Claude thinks, acts, observes tool results, and loops until the task is complete. Features intelligent model routing (haiku/sonnet/opus), 3-layer safety guardrails, SSE streaming UI, and Playwright-powered job auto-apply on Greenhouse and Lever ATS platforms.

• Tech Stack: Python, FastAPI, Anthropic Claude API, Playwright, DuckDuckGo Search, APScheduler, Server-Sent Events
• Features:
  • Manual agentic loop with up to 50 tool-use iterations per request
  • 13 tools: web search, Python execution, file I/O, scheduling, job search, and auto-apply
  • Rules-based model routing — routes each request to haiku, sonnet, or opus in under 1ms
  • 3-layer guardrails: jailbreak detection, tool call sandboxing, and output secret scanning
  • Playwright job auto-apply — fills Greenhouse and Lever forms live in a real browser window
  • Dark-theme streaming web UI with voice input and per-response model tier badges

Description: An ML-powered system for real-time sign language recognition and fall detection. Features MediaPipe Holistic for gesture detection, YOLOv8-pose for human pose estimation, and LSTM networks for temporal sequence classification. Includes data collection tools, Colab training notebooks, and real-time detection with email notifications.

• Tech Stack: Python, Machine Learning, Deep Learning, Computer Vision, MediaPipe, YOLOv8, LSTM
• Features:
  • Real-time sign language recognition using MediaPipe Holistic
  • Fall detection with YOLOv8-pose and LSTM networks
  • Automated email notifications for critical events
  • Data collection and training pipeline with Google Colab integration

Description: Computer Vision & ML automation system for LinkedIn job analysis with OCR, ML predictions, and automated email reports.

• Tech Stack: AI, Machine Learning, Computer Vision, Deep Learning, OCR
• Features:
  • Automated job posting extraction using OCR technology
  • ML-powered job relevance prediction and filtering
  • Automated email reports with job recommendations
  • LinkedIn job board analysis and insights

Description: Advanced computer vision application with facial recognition, emotion analysis, and real-time image processing for intelligent employee attendance tracking using OpenCV, DeepFace, and Flask.

• Tech Stack: Machine Learning, Deep Learning, CNN, OpenCV, DeepFace, Flask, Multi-Modal Emotion Analysis
• Features:
  • Real-time facial recognition for attendance tracking
  • Multi-modal emotion analysis during check-in/check-out
  • Web-based dashboard with Flask backend
  • Automated attendance reports and analytics