Designing a Data Source Discovery App - Part 9: Managing Organizational Data

by DL Keeshin


December 3, 2024



images/future_kds_discovery_erd_20241010.png

Overview

In my last post, I described the design approach for managing location data and using the ChatGPT API to help assign job roles. Today, I am going to discuss the current state of the app design strategy.

The kDS data source discovery app design is evolving. The design now has four key components:

  • Managing organizational data.
  • Generating interviews based on business keys.
  • Distributing and collecting interview data.
  • Analyzing data for better insights.

For this post I'll focus on the first component—Managing Organizational Data—and highlight how aligning data with an organization’s structure is the first step of building a data source discovery interview solution.

Why Align Data with an Organization's Structure?

In most organizations, data sources naturally align with the organization's structure. Each business unit or department collects and manages data based on its responsibilities, needs and function. For example:

  • Sales might use a CRM to track leads and sales performance.
  • Marketing could rely on tools like Google Analytics for engagement data.
  • Finance typically uses ERP systems for budgets and financial reports.

This setup makes sense, but it can also create data silos—where information is isolated in one department and hard to share across the organization. Breaking down these silos helps create a unified view of data, making cross-department insights and decisions much easier.

Why Is This Important?

Aligning your data with how your organization is structured helps with:

Data Stewardship

Data stewardship focuses on managing, maintaining, and ensuring the proper use of data, while ownership belongs to the organization. In recent years, businesses have shifted away from the concept of data ownership by individuals or units, emphasizing stewardship to improve data quality, governance, and collaboration. For instance,

  • HR manages employee records.
  • Sales maintains customer information.
  • Operations handles supply chain data.

Identifying Data Silos

Departments often work independently, which can cause data to be stored in separate systems. This makes it hard to:

  • Get a full view of how the business is doing.
  • Share important information between teams.
  • Make data-driven decisions using all available insights.

Mapping Data Flows

Understanding how data moves across departments helps answer questions like:

  • Where does the data come from?
  • Which teams need this data?
  • How is it used and transformed?

Improving Governance and Compliance

  • Customer data might need to comply with GDPR.
  • Financial data might need to follow SOX regulations.

Managing Organizational Data

1. Collect It

Here are the three main prototype app web forms for collecting organizational data. They are self-explanatory. The first is "Add a Parent Organization", the second "Add A Subsidiary", and the third, "Add Business Units". In this work flow, key attributes like industry, size, revenues, and locations get picked up along the way.

Register Parent Organization
Register Parent Organization
Add Subsidiaries/Divisions
Add Subsidiaries/Divisions
Add Business Units
Add Business Units

2. Assigning Industry Codes (NASIC)

Worth noting -- once a parent organization is registered, the app uses the ChatGPT API to help assign NASIC (North American Standard Industry Classification) codes based on the organization’s description. The python code for generating this NASIC lookup is similar to the script described in my last post that prompts ChatGPT and has it help assign job roles.

NASIC Code Lookup
Register Parent Organization with sMART NASIC Lookup
Why is this beneficial?
  • Improves Accuracy: NASIC codes are often self-assigned by organizations, which can lead to errors or outdated classifications. By using detailed descriptions, the app generates more accurate and relevant industry codes.
  • Easier Data Segmentation: Accurate NASIC codes allow for better segmentation of data, helping organizations align their data sources with the specific needs of different business units.
  • Enhanced Data Quality: Correct industry classification ensures that data is categorized properly, improving the overall quality and reliability of organizational data for analytics and reporting.

3. Linking Data to People

Once the organizational data is gathered, the app needs to know the right person(s) responsible for managing it. Below is the prototype contact form. This data along with organizational details are key items for generating the interview model in the next step of the workflow.

NASIC Code Lookup
DSD Contacts

4. Generating an Interview Model

Having organizational details and contacts in hand, the app will generate interview models tailored for each contact based on the gathered business keys of industry type -function-role. This model was generated using ChatGPT and described in this post.

NASIC Code Lookup
Generated Interview Model

Summary

The kDS app simplifies the management of organizational data by:

  • Centralizing information about organizations and their units.
  • Classifying them with industry-standard codes.
  • Linking data sources to the right people.
  • Improving data governance and discovery.
  • Supporting better, data-driven decisions.

As always, thanks for stopping by.

Leave a Comment: