Project Charter Draft

Business background

• The primary clients are consumer lending companies, banks developing new consumer products.
• Loan acceptance increase, profit increase

Scope

• we are building AI-powered scoring solution provider
• delivery mode: web-based application (probably Microsfot Azure Web App Service #?)

Metrics

Plan

• Phases (milestones), timeline, short description of what we'll do in each phase.

Architecture

• Data

  • What data do we expect? Raw historical data in in xls/csv format from the customer data sources (e.g. on-prem files, SQL, on-prem Hadoop etc.)
  • Sampled data enough for modeling

• What tools and data storage/analytics resources will be used in the solution e.g.,

  • Azure Data Storage
  • Python/TensorFlow for feature construction, aggregation and sampling
  • AzureML for modeling and web service operationalization

• How will the score or operationalized web service(s) (RRS and/or BES) be consumed in the business workflow of the customer?

  • Azure Web App API
  • How will the customer use the model results to make decisions: Power BI connected to Azure Stream Analytics
  • Data movement pipeline in production