There is a disconnect between the data collection needs of field operations and those imposed on them by monitoring and evaluation (M&E) requirements. M&E is often experienced as burdensome for organizations and often results in mismatched incentives that result in delays, biased and low quality results, and worst of all, an inability for organizations to benefit from the very data they collect. Paper processes remain in place while new “solutions” simply add more work to do. The need to digitize data from paper is ubiquitous, despite new Internet connectivity and mobile devices. The key to addressing this situation is operationalizing data collection as part of existing work processes so operations and M&E are both served.

Captricity was born in a Ugandan health clinic and designed to be part of M&E solutions, particularly in resource-constrained contexts. Captricity converts paper images to digital data (including human handwriting) in hours with 99% accuracy. The system algorithmically isolates a form’s individual fields of data, or “shreds,” into distinct images and uses machine learning and computer vision to “read” the data. Captricity combines machines and human intelligence from crowd-sourced workers, who first train the machine learning algorithms, then verify individual shreds – each out of context from the rest of the form so that no one but the client has access to the full dataset. Captricity is HIPAA-compliant and has had government and IRB approvals.

In this talk, we will demonstrate Captricity and discuss ways to incorporate paper-based data into organizational workflows by transforming static data into structured, machine-readable formats for analysis, reporting, and other uses.


Kuang Chen is the founder and CEO of Captricity, a Berkeley, CA-based technology company that converts document images into structured data using human-guided machine learning and computer vision in the cloud. He holds a PhD in Computer Science from UC Berkeley, where his dissertation explored paper digitization from the perspective of data science and human-computer interaction and was the core technology behind Captricity. Prior to graduate school, he was a founding team member of a Seattle-based startup company that developed computational workflow technology for drug discovery laboratories. He is happy to be back at his alma mater, UW, from which he holds a BS in Computer Science and a BA in the Comparative History of Ideas.