Plagiarism detection with competition-proof UX
April – June 2018
Urkund is a plagiarism detection AI used to maintain academic integrity worldwide by educators and their students.
Through our 5-week long Service Vision Sprint, we made sure every voice in the room was heard in the co-creative process where our fresh & highly usable Urkund was conceived.
Designed by engineers, the user experience was either broken or entirely missing, making the old version Urkund difficult to use. Some institutions developed detailed instructions – some as long as 60 pages!
This hugely impacted their customer’s trust in the brand and Urkund’s place among the competition, and Urkund began dragging behind with the rise of strong and more usable competitors worldwide.
The team was small and rigid. We were yet to introduce a new operating model that inspired agility and customer-centricity.
Strong collaboration from the get-go
Starting every week with a workshop, we have developed an understanding of the project goals & vision, shaped new concepts, and decided on key iterations. A consistent approach to time gave everyone peace of mind.
Diverging & converging
We allocated time to thinking in abstraction and making decisions, separating the two distinct mindsets throughout the sessions and activities.
Injecting timely customer insight
We collected a lot of useful insight through problem interviews & concept validation which was later used & synthesized in our co-creative sessions.
Growing a strong design practice through collaboration & injecting customer insight
We stayed true to one fact: design has to be effective, but it also has to be fun! To reach our project goals, we used various methods & tools to get to the desired outcomes & decisions – from Lean Service Creation to more playful LEGO-based methods. As facilitators, we hope that the frameworks were easy to learn & re-apply internally to achieve a long-lasting impact.
Drastically improved usability, navigation, and aesthetic
During customer research phases, we discovered that the use of Urkund differed throughout geographies. The prime difference was in time allocation used to interact with the tool.
As a result, we designed UX flows that reflected key goals of the users for higher levels of usability, enabling various ways to interact with Urkund.
The new screens and UI elements asked for expansion of colour palette and Urkund identity to enable the designed flows to come true through solid navigation, contrast, and consistency.
Improved sense of trust in the AI
Urkund results are rigorously trustworthy. The educators in certain geographies would set the minimum required score to assess the work of their students, which resulted in numerous student works marked failed. In a few worst-case scenarios, some students lost their study places! Shocking, but true.
We encouraged giving transparency to the Urkund scores in order to encourage educators to trust, but carefully. After all, AI, too, makes mistakes, and those mistakes shouldn’t decide the student’s fate.
In September 2019, Urkund was released to the public.
A year later, Urkund and PlagScan joined forces as Ouriginal, an ultimate plagiarism detection add-on to ensure academic integrity at scale.
Urkund was developed during my gig at Futurice. Hence the project deserves to be attributed to all parties.
Anastasiia Kozina (me)