We look forward to receiving your signal!
Inpher: AI-assisted data analysis that preserves privacy
Data analysis and privacy needn’t be incompatible. Inpher’s analyst solution Secret Computing® leverages exactly that. We assisted in the conceptualization and design of the portal for data analysts.
In a nutshell
- The portal’s complex logic required us to balance technical requirements against design requirements. We had to understand the underlying machine learning processes to design an intuitive and easy-to-use interface.
- Our first minimized version served Inpher as a proof-of-concept with their customers.
- Through our long-term close collaboration with Inpher, we are able to continuously and incrementally improve the portal.
The data and the portal
Confidentiality is a main concern in many sectors such as finance, health and insurance. The big challenge in data analysis is the need to move and use data collected in different sectors without violating confidentiality agreements or sovereign data privacy policies. The Cambridge Analytica scandal showed how this should not be done. But how can data research be enabled without breaching the users’ privacy?
Inpher is a Swiss start up offering the solution to said problem: GDPR-compliant machine learning. Their technology powers advanced analytics and AI applications without exposing or transferring sensitive data across organizations or jurisdictions. Like this, a lot of valuable data is made available to data scientists.
We have been working with Inpher on conceptually redesigning and improving the interface of their analyst portal while simplifying it for users. Our design makes the maintenance of the portal straightforward, while supporting functionality for entry-level to advanced data scientists.
Prototyping, design and the first interface
Our work was characterised by rapid iteration cycles with continuous feedback from our client. Through this a very lean approach was possible:
1. To begin, we collaboratively developed high level prototype personas. This helped to outline and align on who this portal for.
2. Then we started to design the interface based on Inpher’s existing (prototype) portal. This allowed us to understand the technical challenges that the portal has to cope with. We led Inpher through multiple rounds of ideation and, using pen-and-paper wireframes, designed the new interface.
3. After the alignment with Inpher, we developed the first, simplified version of the portal using state of the art Vue.js. Inpher used this first version as a proof-of-concept with clients to get feedback. We incorporated feedback right away and pushed for continuous iteration. The long-term collaboration gave us the opportunity to test, improve and iterate.
Since the portal’s release, we have kept working with Inpher on improving the interface and the user experience in general. The knowledge we acquired about the portal allows us to efficiently and constantly refine the interface through the integration of new functionalities and simplification of others.
We are proud of this successful collaboration built on mutual trust. We’re also happy to see our work with Inpher strengthen our long-term partnership, and we are excited to see where this goes next.
Questions? More about our work: contact Bengiamin Barblan (firstname.lastname@example.org).