Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  • Roadmap and release updates
  • Certification changes and governance process
  • Identification code feature for the API
  • Data Standard feedback
  • "Experimental" Flag
  • Ed-Fi Summit reminder

Materials

View file
nameTechnical Advisory Group August 2023.pdf
height250

Participants

Expand


First NameLast NameOrganization
JoshBergmanSkyward
DirkBradleyMichigan Datahub
DavidClementsEd-Fi Alliance
WyattCothranSouth Carolina Department of Education
KatieFavaraTexas Region 4
StephenFuquaEd-Fi Alliance
Jean-FrancoisGuertinEdWire
JasonHoekstraEd-Fi Alliance
MattHoffmanAeries Software
ManuelJaramilloTexas Education Agency
SherodKeenKeen Logic
VinayaMayyaEd-Fi Alliance
NaduNairWalla Walla Pubic Schools
RonPeashaInfinite Campus
AndrewRiceEducation Analytics
LucySauraLake Washington School District
RickThompsonSouth Carolina Department of Education
MaureenWentworthEd-Fi Alliance
MustafaYilmazEd-Fi Alliance

Support: Ann Su

Notes

These notes summarize some of the discussion supplemental to the details already in the slides above.

  • Summit registration
  • Identification code
    • Continues to be a pain point for many implementations:
      • A SIS may have many different identification codes for each student.
      • An LEA may send different identification codes to different systems, on accident or for some legitimate reason.
      • An SEA may use state code when setting up state-level assessments, while LEA is using their local code for other assessments.
    • Current solutions
      • Some are still using ODS/API v2.x for its identification code lookup when receiving assessments.
      • Custom ETL integrations that pull source system data (or files), transform, and then load into the Ed-Fi API.
      • Data Import scripting.
      • ... or just stuck and cannot integrate.
    • Potential solutions
      • Add more lookup querying / filtering to the Ed-Fi API. Pro: relatively easy to use, and scalable from a host perspective. Con: very chatty process.
      • Restore the translation capability that was in v2.x. Pro: translates on the fly. Con: performance hit; currently relies on caching that doesn't scale well.
        • Facilitates faster on-boarding of new providers instead of getting stuck with workarounds or failure to be able to integrate data.
    • Alternatives? Perhaps provide more options for rostering into providers who haven't integrated with Ed-Fi
        • More control over which identification code is used.
        • OneRoster?
    • Ed-Fi commitment to continuing the research on questions above and get back with TAG and the community.
  • Data Standard "Experimental flag"
    • Must be very clearly marked as such in the API
      • Swagger / Open API extended attribute
      • Stephen's note after the meeting: perhaps even with the naming, like having an X at the beginning of the name
    • If present, would like to have a feature flag for disabling experimental data standard attributes and domains.
    • Response generally seemed receptive.
    • Need to define in more detail as proposal for 2024.

Next Meeting

members: please send any agenda suggestions.

Action Items

  •  Stephen Fuqua write up additional details on proposed "Experimental" flag in the Data Standard
  •  Vinaya Mayya (and others as needed) continue investigating options and developing a specific design proposal for identification code lookup
  •  Stephen Fuqua internal Ed-Fi discussion around potential for pulling well-defined roster files from the API