A live selection of my work in machine learning, AI, and applied software — pulled straight from GitHub and always up to date.
Deep-learning system that fine-tunes a VGG-16 CNN (transfer learning) to classify structural crack severity — Hairline, Moderate, Critical — with OpenCV contour analysis to estimate crack dimensions and a Streamlit app for real-time inference.
Submitted a test-suite optimization to the OpenML Python library: mocked server-side API responses (session.mock) to convert network-dependent tests into hermetic, deterministic unit tests — cutting module runtime ~87% (140s 18s).
Speech-to-speech translation pipeline that transcribes audio, translates the text, and synthesizes it back to speech.
Model and training code powering the Audio Translator speech-translation pipeline.
Real-time telemetry and monitoring dashboard for precious-metals data, built with Grafana.
Self-hosted file storage server with a clean upload/retrieval API.
A news website that refreshes its content automatically from live sources.
Web-scraping dashboard for collecting and visualizing data from the web.
Auto-synced from github.com/Sachin0496 — new repositories appear here automatically.