Case Study: Noitso achieves model deployment in hours instead of days with Seldon Deploy

Noitso accelerates model deployment from days to hours

Noitso, a Copenhagen-based data science firm that builds credit ratings, scorecards and risk profiles, faced slow, error-prone model deployments and limited explainability and monitoring using their previous C-based workflow. To modernize their stack and develop in Python with container-native tooling, Noitso turned to Seldon and adopted Seldon Deploy to accelerate and standardize production delivery.

Using Seldon Deploy (with monitoring via Alibi Detect), Seldon enabled Noitso to cut their model review and deployment cycle from multiple days to just hours, improve production monitoring and detect when retraining is needed, and roll the faster, more explainable models out to more customers. The result: quicker proof-of-concept validation, greater trust in production models, and a measurable competitive advantage for Noitso.

Noitso

Thor Larsen

Data Scientist