Case Study: deepsense.ai tracks 120K+ ML models efficiently with Neptune.ai

How deepsense.ai Tracked and Analyzed 120K+ Models Using Neptune

deepsense.ai, an AI-focused software services company, needed a way to manage a very large machine learning project for a leading CEE food company. The team had to analyze the impact of promotional campaigns on sales across thousands of subproblems, ultimately training more than 120,000 models without an existing experiment tracking solution, which made tracking, comparing, and collaborating difficult.

Using Neptune.ai, deepsense.ai implemented experiment tracking, metadata logging, dashboards, and seamless Optuna integration to organize and compare models at scale. Neptune.ai helped the team store visualizations, predictions, and pickled models, track 120K+ experiments, and save weeks of manual work while improving productivity and decision-making for the sales-impact analysis.

deepsense.ai

Patryk Miziuła

Senior Data Scientist