Specifications for HUD HMIS Point In Time count app support - servinglynk/hslynk-open-source-docs GitHub Wiki

Minimum viable app support features required

• Full-refresh upload of data set from app, on behalf of a specific PIT count client community/CoC.

• Data warehouse access to uploaded data

• Accommodate additional PIT data elements which are not within the HMIS Data Standard APIs, using one of two approaches already implemented:

  • Relational: population of all required data elements into existing or additional relational schemas (implemented in HSLynk as PostgreSQL schema)
    • Custom survey questions can be stored in the survey APIs, which store question and response and can optionally map to HUD HMIS data elements.
  • Graph/Data Lake: single, full refresh upload support

• Trusted App registration and OAuth2 authentication

Potential app support features

• incremental record uploads by PIT app

• editing previously uploaded records, from the PIT app

• canned PIT real-time dashboards from the warehouse

Question: Do all PIT data collection implementations use anonymous clientele, or are obfuscated or identified client records ever used?