One Month to Make a Difference: Automating Affordable Housing Applications

September 17, 2015 7:00 am PST | Data as a Service, Data Rockstars
Photo: Kjetil Ree

Marcus Louie participated in Socrata’s “One Month to Make a Difference” Program, a paid sabbatical to implement an open data project.

In the first installment, Marcus learns about the time-consuming, PDF-based application process for Section 8 housing in the District.

Next, together with a Code for America project team and the non-profit Bread for the City (Bread), he helps create a solution for automating the completion of multiple applications that saves residents time in applying for low-income housing.

Start With Priorities

Stacey Johnson, a Senior Case Manager at Bread for the City, and I prioritized my work on this project as follows:

  1. Improve work flow. It should be more consistent with how their caseworkers actually process new and existing clients.
  2. Begin an integration between Salesforce. Bread uses Salesforce CRM to hold all their client data, so we’d need to pull data out of Salesforce to drive the application we’re building. At the same time, Bread only just recently finished modeling a lot of the data that our application processes — income history, rental history, criminal history, and household member information. That means they don’t actually have any clients with their information stored in Salesforce.
  3. Identify data fields. Determine the data fields that help quantify and provide context for the demand for Section 8 housing in the District.

Andrew Lomax, who manages Bread for the City’s Salesforce instance, walked me through the data model changes they’re putting in place. We were previously pulling from a single Salesforce object, but they have now expanded the number of fields they’re collecting about a client and spread them across multiple objects.

After this discovery, I documented all of the fields that we were capturing about a person and mapped them to the fields that Bread for the City was capturing in Salesforce. Our initial Salesforce integration was minimal and based upon the previous Salesforce model: first and last name, DOB, mailing address, and SSN. With the new Salesforce model, we would need to map all of these new Salesforce fields and then define the gap between Salesforce and our application.  

mapping documents between Salesforce and a District Housing application
Screenshot of the mapping between our application, “District Housing,” and Bread for the City’s Salesforce instance.

Up to this point, we were testing our application with a handful of fake clients, enough to list on a single screen. In reality, Bread has helped more than 2,000 people through their Housing Access Program and they have another 90,000 people in their client database, so a caseworker could potentially perform a massive search. In our new architecture, caseworkers can type in the name of the client, and the list of clients is progressively filtered down as they type.

Caseworker interactions with Section 8 applicants' information.
Screenshot of one of the interactions I worked on in the first week. Caseworkers can filter the list of clients by typing a client’s name. The list is progressively filtered as the caseworker types.

Deployment and Testing

I spent the last week of my sabbatical visiting Bread for the City’s offices. We are close to making affordable housing just a little bit easier in Washington, D.C. Since then, we have pushed the remaining changes to our GitHub repo and Bread for the City has deployed our application onto a dedicated machine to run from within their network. This was a big step for us as it meant that are now able to test the application with caseworkers and then eventually with real clients.

The project is currently in the final stages of testing and here are the next steps to implementation:

  • Technical Implementation: We deployed the application inside Bread for the City’s network and are wrapping up the testing stage. The next step is to start to use the application with a small number of clients first before rolling out to use in their weekly Housing Access Program workshops.
  • Full rollout: When the app is in use, it will be used twice a week during a housing workshop run by Bread for the City.
  • Opening the data: The Urban Institute and the D.C. Deputy Mayor’s Economic Intelligence team have been participating on how to release affordable housing data in the District as well as on a data model that can be used to provide insight into the demand for project-based Section 8 housing.

Solving the problem longer-term: We caught the attention of the U.S. Department of Housing and Urban Development. We’ve talked with the Deputy Secretary’s office and and are scheduling a demo to discuss the project and how the problem could potentially be solved at a higher level. The ultimate goal is to make this project completely irrelevant by having it solved by HUD. If you want to help emphasize the importance of this work, you can vote for it on HUD’s idea board.

To find out more about how Socrata can help you achieve your top initiatives, schedule a personalized demo of our industry-leading solutions.


Request a demo


Previous Article
Civic Awesome, Data News Roundup
Civic Awesome: Open Data in the News for the Week of September 14, 2015

September 18, 2015

Next ArticleThe wetlands in Alastka's Bristol Bay Watershed
Data as a Service
Open Data Power: Saving the Planet

September 16, 2015