Deepiction
Making your own data readable
A digital companion through the confusing world of your own data: Deepiction translates everything that piles up about a person across every platform into one complete, clear overview — and hands back control over it.
Anyone who requests their data from a platform only gets to it after a long detour through the privacy settings, and then receives a list with no structure, in a different format from every provider. Deepiction starts there: you ask your own data a question, such as where you spent the last few days or which party you will vote for in the next federal election, and see what can be read about you from it. The project ran in the sixth semester, five of us across two universities: three designers at the HfG and two developers at Aalen University, who built the search engine as a working prototype.
Role
Context
March – July 2021
Supervision
Everyone generates data. Almost nobody can read it.
Problem
The GDPR gives everyone the right to ask a platform for their data. What comes back is a raw export — thousands of lines of JSON, unstructured and unexplained. The right exists; the understanding does not.
- No real sense of security about what happens to personal data
- Almost no education about how that data is processed
- Filter bubbles quietly remove both the overview and the choice
- The legal right to your own data returns files nobody can read
Research
We ran our own survey well beyond the school's bubble and interviewed users alongside three experts. Several participants requested their real exports from the platforms and let us work with them — which is how we learned first-hand how unreadable those files are. Two personas came out of the testing, a Hardliner and a Softliner; designing for both is what kept the tool from turning into a privacy lecture.
Expert interviews
- Andrew HillFounder, Textile Photos — decentralised data infrastructure
- Marcus GelderieProfessor of internet security, Aalen University
- Michiel Top & Vivien RoggeroCEO and CTO, Your Digital Self — a startup building almost exactly this
Concept
We built and tested six directions rather than arguing about them — a timeline, a data visualisation, data management, a browser plugin, an open question box and an avatar. A weighted matrix settled it: the avatar drew the most interest, data management answered the most need. Instead of picking a winner we kept the strongest part of each, and built against two questions.
Two questions to build against
What has been said about me online?
Which party would I vote for in the next federal election, judging by my data?
The second is the sharper of the two. It takes a single sentence to show how far everyday data can be read.
Approach
Non-profit, funded by donations — no data sales, no advertising. Four values sit on the login screen: autonomy, transparency, responsibility, security. The interface is deliberately calm and explanatory rather than alarmed.
A translator, not another platform.

Registration

Search


Result pages

Privacy settings
The overview
Explaining itself
Q&A pages explaining the tool, the organisation behind it and the data world it operates in.
It ran, and it was tested.
Two teammates at Aalen University built a Node.js backend with a full-text search over a synthetic dataset drawn from ten sources, designed to answer exactly those two questions. My part was the interface and the translation between design and build. We then ran a usability lab session in front of about twenty people; I moderated.
What came back
- The registration flow built too little trust, and its steps looked too alike
- Posting a master password by mail read as old-fashioned
- The filters needed far more visual weight to be found at all
The teaser in full
Team
- Hana Amer · HfG Schwäbisch Gmünd
- Jasmin Klement · HfG Schwäbisch Gmünd
- Alina Schadt · HfG Schwäbisch Gmünd
- Anjan Bösinger · Aalen University
- Tobias Wörnle · Aalen University
Credits
Supervised by Prof. Hans Krämer · “User Driven (Product) Design Development”, summer semester 2021, in cooperation with Aalen University.
Outcome
What we learned, including the awkward part.
Halfway through the research we found Your Digital Self — a startup building a data wallet on almost exactly our premise. We interviewed their CEO and CTO rather than quietly changing course; they offered a collaboration, which we turned down for the semester and left open for afterwards. Finding out that your idea already exists, and talking to the people who got there first, taught us more than another round of sketching would have.