peteroyce - Overview

Pinned Loading

  1. Multi-modal semantic search engine powered by CLIP embeddings. Index images and text, search across both with natural language.

    Python

  2. Document intelligence pipeline. Extract, classify, and analyze content from PDFs, images, and scanned documents with OCR, layout analysis, and NER.

    Python

  3. ML model monitoring toolkit. Detect data drift, model degradation, and prediction anomalies in production with automated alerts and dashboards.

    Python

  4. LLM evaluation and comparison framework. Benchmark models across tasks with automated scoring, human preference tracking, and detailed analytics.

    Python

  5. Predicts nonprofit risk of losing exempt status. Trained on 1.8M IRS records

    Python

  6. real-time anomaly detection engine for time-series data

    Python