Data Science Toolkit

Open-source, curated collection of proven machine learning implementation accelerators, automating commonly repeated development processes. Allowing data science practicioner to focus more time on complex business value.

Repos:

Machine learning

• Anomaly Detection Solution - Detect anomalies on very large structured data sets
• Binary Classification - Binary classification, with parameter based auto algorithm selection
• Forecasting Solution - Pre-configured engine for demand forecasting, map data into the existing model to generate a forecast
• Hierarchical Multilabel Classification - Apply classification to hierarchical structured classes
• Large-Scale Route Optimization Accelerator - Solve large-scale route optimization where the optimal solution cannot be found in reasonable time
• Retail Analytics - Customer Segmentation, Churn and Lifetime Value Prediction

Conversational AI

• Conversational AI Advanced Pre-processing Service - Collection of modules to help with validation, identification and authentication processes
• GLUE - Azure Cognitive Services - GLUE is a lightweight, Python-based collection of scripts to support you at succeeding with speech and text
• Power Virtual Agent AudioCodes Contact Centre accelerator - Horizontal solution that addresses cross-industry Contact Center needs
• Speech-to-Text Transcription and Classifition Solution - E2E speech transcription and message classification solution

Knowledge mining

• Knowledge Graph Powered Search - Implement a domain-aware search solution using knowlege graph (KG) to enrich the organic results
• Knowledge Mining - Use intelligent services to quickly learn from vast amounts of information
• Knowledge Mining (Verseagility) - Knowledge mining on unstructured data sets with no data labeling

ML Ops

• ML Ops - Configurable CI/CD pipelines, AML pipelines, and compute resources for ML Ops
• ML Ops for Databricks - ML Ops support for Databricks

Computer vision

• Azure Object Detection - Collection of pipelines to start a template object detection project with as little effort as possible
• Object Detection - Uses computer vision for object or defect detection and includes edge deployment capabilities

UI

• Vitastic - Quickly build web-interfaces for object detection, segmentation and classification models

Responsible AI