Case Study: Fortune 500 Telecom Company achieves +25% accuracy and trains 200k labels in hours with Snorkel AI

Fortune 500 Telecom - Customer Case Study

Fortune 500 Telecom Company engaged Snorkel AI and used Snorkel Flow to classify encrypted network data flows into their associated application categories. The customer’s challenge was a shortage of labeled training data—creating labels was slow, required deep network expertise, and blocked deployment of AI-enabled network applications.

Snorkel AI implemented programmatic labeling with Snorkel Flow to leverage unlabeled and partially labeled traffic and to enable comprehensive network data exploration and analysis. The telco trained about 200k labels in hours, achieved +25% accuracy above their ground-truth baseline, and saw a +75% accuracy improvement on a critical data slice.