Validation Research - GSA-TTS/document-extractor-poc GitHub Wiki

Validating Specific Problems

The scope of our PoC was grounded by a three-month phase of validation research focused on developing a user-centered understanding of the most salient problems within the document submission space.

To kick-start this research, we created visual mock-ups to gather feedback from potential users on an array of product concepts. After understanding what components of a future solution were most desirable, we scanned the government-technology vendor landscape to understand available market solutions and identify what document processing gaps exist.

Some of the key insights we uncovered:

  • Applications and document collection methods are increasingly becoming mobile friendly for members of the public, allowing more flexibility and choice to submit necessary documentation digitally.
  • Increased flexibility for members of the public shifts burden to administrators, with many staff hours spent on coding static files (images, PDFs) into agency data systems.
  • Agency procurement approaches for IT modernization are often large, complex, multi-year engagements. While document processing technology, like optical character recognition (OCR), is not new, such procurement approaches often lead to the continued use of outdated technology.

Recommendation

Based on our research, the team validated that a document extraction solution was well-positioned to meet TTS's criteria for success, delivering impactful, scalable improvements to document processing across government agencies.

  • Desirable: Nearly all our interviews with potential customers (state administering agencies), delivery partners in the document solution space, as well as data management system experts, have pointed to a need for better document processing tools for public sector agencies.
  • Viable: Reducing manual administrative burden is essential for agencies operating with tight budgets and limited staff. Business departments are willing to invest in tools that enhance efficiency and have historically trusted federal solutions.
  • Feasible: Our proposed solution aligns with existing commercial technologies. While OCR and similar technologies are not new, recent advancements make them a strong foundation for testing and future growth.
  • Valuable: A “Better Document Submission” tool has wide applicability; any agency that collects information from its customers has a potential use case. This is an ideal case for TTS’s ability to work across government to make it easier and cheaper for people with the same challenges to use the same tools.

Research deliverables that informed our recommendation to move to prototyping:

  • Recommendation Deck outlines findings from our validation phase and recommendation to move forward to prototyping.
  • Customer Insights Deck - A research summary of key insights gathered from talking to potential customers (e.g. government agencies) about their needs and pain points related to document submission.
  • Existing Solutions: Product Comparison (xlsx) - A market analysis of existing products in the document submission space.
  • Potential Use Cases (xlsx) - An exploration of typical documents requested across public benefit programs and what key data is manually extracted.