Sleep & Dream Database: Searchable Library of Dream Data 

The Overall Problem

As an authority in the field of dream research, Kelly Bulkely, Ph.D. has long been gathering and curating dream data in an effort to provide a wide-ranging data set to fellow researchers and the general public alike. Not only does this data set include over 33,000 long-form dream journal entries, but also includes both qualitative and quantitative data from over 14,000 survey respondents about their dream habits, including demographic information about the respondent such as age, education level, and political affiliation.

Though well-intentioned, the original iteration of the application suffers from poor performance, a dated UI (including no feasible mobile experience), and a frustrating UX that often brought users to unexpected or even empty results pages. Additionally, continuing to update the application with new and salient data in this iteration is an overly convoluted and tedious process.

The new version of the application provides the user with a predictable experience across devices of any size, guiding them towards robust results with visual cues that help highlight key statistics or potential correlations within their search. In addition, the process of adding new data to the application will be streamlined to allow newly-gathered survey responses and dream journal reports to be quickly available to fellow researchers.

Our Solution

1. Understanding the user

Different users expect different experiences, but dictating that experience can feel restrictive and make the application seem disjointed. By understanding core user journeys and planning accordingly, the GRAYBOX team was able to create a seamless application that could gracefully cater to users of any experience level in regards to navigating dream data. GRAYBOX also understands that even for data-driven applications, a first-class mobile experience is a must, and mobile/tablet users were considered and planned for in all phases of the design process.

2. Re-architecting the data

Data-driven applications must use tools that meet the needs of the data. In this case, the ability to quickly search over hundreds of thousands of words quickly and accurately is of tantamount importance. Additionally, we must allow the size of the data to grow without worry of leaving the tooling behind. By utilizing Elasticsearch for the dream journal searching while relying on a GraphQL schema to interchange data between the server and client, the GRAYBOX team achieved a data architecture capable of lightning-fast searches and virtually limitless scalability.

3. Partnering for success

A partnership works best when both parties come together with their own unique understanding and area of expertise and find a way to unite. By striving to fully grasp how dream research is performed, the GRAYBOX team worked collaboratively with SDDb, coupling our expertise in the digital space with SDDB’s domain knowledge of dream analysis to deliver the best possible end product.