Health & Human Services

PROJECT OVERVIEW

GrantSolutions processes over $750 billion in Federal Grants yearly. There was a heavy administrative burden which led to slow manual processing of federal grant applications and siloed data difficulties. GS was also working on antiquated legacy systems which needed updating.

With a design thinking and user centered approach, we began by conducting interviews with financial experts and federal tax auditors working at the Health and Human Services and the Administration for Children and Families offices. We also co-designed with grantees and subject matter experts to fully grasp the experience of all stakeholders involved and developed personas to use as guidelines.

With the information gathered from our interviews, I was able to lead the Machine Learning team in ideation sessions for potential automated tools and services that would add real value to a wide range of users from the beginner to the super user.

As this was a complex problem with many moving parts, we developed a multi-pronged approach including natural language processing of applications and condensing of FAC data to reduce 30-40 page reports to a single page of pertinent data. We also were able to aggregate 30 years of grant spending data to identify unusual spending habits of grantees while tracking their individual spending data to make predictions on future spending.

After getting the green light from the executive team, I then created wireframes, and mock-ups which we would present to the executives on a weekly basis. These solutions were then tested with focus groups at the HHS ensuring that our UI was visually intuitive.

Throughout the process, we ensured that our machine learning tools were truly adding value to the most evident pain-points of both the auditors and the grantees. These were just tools that helped to expedite the grant making process. To reinforce this idea we provided complete transparency of potential risks and errors from the automated data sets by providing tool tips that would provide information into how the AI consolidated information and why.
‍
At the end of a 12 week period, we successfully implemented intuitive search and rapid consolidation of data with automated intelligence that was fast and scalable within the cloud while sticking to a cost effective rapid build.
With our help, GrantSolutions was able to increase their efficiency from weeks to seconds while saving an estimated $30 million of tax payer money per year. GrantSolutions Recipient Insight is now able to have swift communication and access to over 750,000 organizations and counting.

Thank you for reading. Check out some of my other projects below.