top of page
  • rszczurek

Can You Help Me?


How has F33 used Vertex AI to optimize help desk processes at Inlogica company?

Customer challenge

Our customer, Inlogica company has achieved substantial success in its marketing activities by selling Microsoft Dynamics services. This increased the number of serviced customers and, as a result, the number of service desk tasks to be handled by the team. The Inlogica team observed that around 70-80% of tasks are very similar and answering them could be automated. Moreover, handling the next 20-30%, the more complex ones, can be also optimized by providing the consultants with a set of possible questions to choose from.

Our solution

F33 has followed its AI/ML Framework to deliver the AI/ML solution for Inlogica. As a result we helped Inlogica to formulate requirements, prepare datasets (using their history of contacts with customers) and upgrade their IT infrastructure to be able to communicate with the model.

For the AI/ML model, we decided to try the Vertex AI AutoML NLP as a benchmark model. For the next stage, we will evaluate other approaches like the BERT language model using Tensorflow TFX framework.

Models were implemented using F33 MLOps GCP Platform and the diagram below presents all core components used for this implementation.


During the business understanding stage we identified that customer wanted to optimize two metrics:

● Reduced human intervention in repetitive tasks by at least 50%.

● Increased throughput in non-repetitive tasks by at least 20%.

21 views0 comments

Recent Posts

See All
bottom of page