• Products
    • Observe

      Keep tabs on your world in real-time, reducing a cacophony of activity from massive amounts of raw content into orderly, easy to consume data

    • Orient

      Our low-code, composable, distributed, and event-driven predictive analytics workflow engine mines for actionable insights hidden in data at any scale

    • Dominate

      This insight interaction and decision support application takes you beyond self-service, visual exploration of piles of insights & provides auto-ML powered actionable, scenario-specific advice to achieve the goals you specify

  • Markets
    • Government

      At BigBear.ai, we take pride in providing mission-critical services and solutions to our government customers. This includes Data Analytics, Systems Engineering, and Cyber services.

    • Commercial Markets

      BigBear.ai provides innovative AI solutions to companies across commercial markets, ranging from Space to Media to Shipping and Transportation.

  • Company
    Just Announced
    Julie Peffer and Josh Kinley
    Press Release
    BigBear.ai Announces Key Leadership Appointments in Finance and Corporate Development
    • Col – 1
      • About

        We help governments and businesses make the decisions that change markets and define outcomes

      • Investor Relations

        We operationalize artificial intelligence and machine learning at scale through our end-to-end platform

      • Partners

        We work with our technology partners to customize products for the private and public sectors

    • Col – 2
      • Team

        Our executive team brings decades of world-class experience

      • Newsroom

        News articles from our press room

  • Careers
    Apply Now
    Data Scientist/Analyst
    Location:

    San Diego, California

    Description:

    BigBear.ai an exciting opportunity for a Data Scientist/Analyst to join our team in San Diego. This position can be onsite in San Diego, CA office or remote.

    Category: Computer/Data Science

    • Col – 1
      • Explore Jobs
        • Returning Applicant Login
        • Current Employee Login
      • Culture
      • Benefits
      • Military and Veterans
      • Product and Technology
      • Join Our Talent Community

        Sign up to receive personalized alerts and stay up to date on job openings right for you

  • Resources
    Fostering AI Adoption - Dr. Reggie Brothers - GovCon DIFFERENT Podcast
    Dr. Reggie Brothers
    Video
    Fostering AI Adoption - Dr. Reggie Brothers - GovCon DIFFERENT Podcast
    • Col – 1
      • Blog

        See what’s new in artificial intelligence, machine learning, and data analytics

      • Resource Library

        Browse our resource library and discover more about our products and solutions

      • Newsroom

        News articles from our press room

  • Contact
BigBear.ai
  • Products
    • Observe

      Keep tabs on your world in real-time, reducing a cacophony of activity from massive amounts of raw content into orderly, easy to consume data

    • Orient

      Our low-code, composable, distributed, and event-driven predictive analytics workflow engine mines for actionable insights hidden in data at any scale

    • Dominate

      This insight interaction and decision support application takes you beyond self-service, visual exploration of piles of insights & provides auto-ML powered actionable, scenario-specific advice to achieve the goals you specify

  • Markets
    • Government

      At BigBear.ai, we take pride in providing mission-critical services and solutions to our government customers. This includes Data Analytics, Systems Engineering, and Cyber services.

    • Commercial Markets

      BigBear.ai provides innovative AI solutions to companies across commercial markets, ranging from Space to Media to Shipping and Transportation.

  • Company
    Just Announced
    Julie Peffer and Josh Kinley
    Press Release
    BigBear.ai Announces Key Leadership Appointments in Finance and Corporate Development
    • Col – 1
      • About

        We help governments and businesses make the decisions that change markets and define outcomes

      • Investor Relations

        We operationalize artificial intelligence and machine learning at scale through our end-to-end platform

      • Partners

        We work with our technology partners to customize products for the private and public sectors

    • Col – 2
      • Team

        Our executive team brings decades of world-class experience

      • Newsroom

        News articles from our press room

  • Careers
    Apply Now
    Data Scientist/Analyst
    Location:

    San Diego, California

    Description:

    BigBear.ai an exciting opportunity for a Data Scientist/Analyst to join our team in San Diego. This position can be onsite in San Diego, CA office or remote.

    Category: Computer/Data Science

    • Col – 1
      • Explore Jobs
        • Returning Applicant Login
        • Current Employee Login
      • Culture
      • Benefits
      • Military and Veterans
      • Product and Technology
      • Join Our Talent Community

        Sign up to receive personalized alerts and stay up to date on job openings right for you

  • Resources
    Fostering AI Adoption - Dr. Reggie Brothers - GovCon DIFFERENT Podcast
    Dr. Reggie Brothers
    Video
    Fostering AI Adoption - Dr. Reggie Brothers - GovCon DIFFERENT Podcast
    • Col – 1
      • Blog

        See what’s new in artificial intelligence, machine learning, and data analytics

      • Resource Library

        Browse our resource library and discover more about our products and solutions

      • Newsroom

        News articles from our press room

  • Contact
Home Blog Designing Operational Data Stores
Blog

Designing Operational Data Stores

Jim McHugh
February 5, 2020
  • Share
  • Share

Operational Data Stores (ODS) are entity-centric and should only contain data about that specific entity. When designing an ODS, you will use relational database design methodology. I try to match my ODS data model as closely as possible to the source systems data model. By keeping the models similar, you will make the loading and synchronization of data to and from ODS and the source systems easy and efficient.

Let’s take a look at a Constituent (Person) for a membership and publishing company. If you remember from my previous post on how to use an ODS, this company has many different personals a single person can have when interacting with this company. They can be an author, an editor, a reviewer, a member, a student, an educator, a meeting attendee, a customer, or on online contributor, etc. How would I design an ODS for such a wide range of personas?

For the Constituent ODS, I only care about all of the information concerning the Constituent. This is more than name and address. I care about what manuscripts they have submitted, what articles were published, what journal they review for, etc. This information is critical to the operational usage of the ODS. What I don’t care about is what school they were teaching at in 2015 when they published an article or what did the previous version of this manuscript look like. History does not live in the ODS; that information is stored in the data warehouse.

Consider a person who is a member, author, editor, reviewer in the peer-review system, a subscriber to one or more journals in the journal delivery system, and an active member of the online community application. As you can see, this single person has many personals and may, in turn, have many distinct values for key information that will need to be addressed so we can have a complete view of this person.

For this person, we can use the member number as the unique identifier and create a single record for this person with flags denoting the different roles this person plays in the various systems. For authors, reviewers and member s of the online community application who are not members we can try to combine the data using a unique attribute like email address or cell phone number. Again, we would use flags to define the roles this person plays. Now I can easily ask how many authors are members?

Here is what an ODS data model might look like for persons interacting with this organization:


As you can see there is a single table containing person data. Using flags for the roles a person plays when interacting with the organization provides great flexibility in answering the operational questions an ODS is created to answer. This design will also allow me to synchronize the changed data with my source systems.

I hope you see the power a properly designed ODS can provide your organization.

BigBear.ai

2022 BigBear.ai • All Rights Reserved.

  • Privacy Policy
Corporate Headquarters

6811 Benjamin Franklin Drive, Suite 200
Columbia, MD 21046
Tel: 410.312.0885 • Email: [email protected]

CareersContact