1. Piktiv
  2. /
  3. Insights
  4. /
  5. Case
  6. /
  7. Minna technologies

Minna Technologies - Powering the global subcription economy

Background

Minna Technologies is a Swedish fintech company that provides subscription management solutions through bank integrations. Due to the large volume of users, back office staff and data input into the system was significant.

MinaLogoNew

Solution

One part of the solution allows end-users to cancel ongoing subscriptions easily. Minna, with the assistance of end-user request determination, automates the request and supplier response. The supplier response contains relevant information such as the notice period, termination date, and possible binding time. Two different models, a convolutional neural net, and a recurrent neural network, were used to extract the necessary information.

Results

Piktiv developed a deep learning NLP (natural language processing) model that accurately identifies the meaning of the supplier responses and extracts relevant information that should be sent to the end-user. This automated response allowed Minna to automate most supplier responses, making their solution infinitely scalable.

Programs

Python, Tensorflow, Keras, NLP, Convolutional Neural Network, Recurrent Neural Network

Customer case

image (3)
Case

Impulso

Valheim game characters with iron gate and piktiv shields
Case

Valheim

videquus1
Case

Videquus

Piktiv – Bild
Case

Brand New Profile

Begagnathall minimize
Case

Mercari Group