Case Study: NTT achieves real-time, scalable streaming flow analytics (100K+ events/sec) with Imply (Druid & Pivot)

How NTT Powers Their Analytics Stack and Data Exploration with Imply

NTT, a global telecommunications company (NTT Communications / Global IP Network), needed to modernize its legacy flow‑analytics stack that had become a black box, didn’t scale cost‑effectively, and provided limited ad‑hoc analysis for its traffic‑matrix use cases. After evaluating options, NTT built a Kappa streaming architecture using Apache Kafka for ingestion and Apache Druid for queries, and worked with Imply to leverage Druid and Imply’s Pivot UI for ad‑hoc exploration and visualization.

The solution implemented a Kafka→Druid pipeline with Imply’s Pivot for analytics, enabling end‑to‑end streaming analytics that ingests more than 100K events/sec in production. With Imply, NTT gained scalable, efficient query performance (fewer resources than alternatives), fast ad‑hoc GROUP BY/WHERE analysis across dimensions, and cross‑organization access to create and share custom dashboards — unlocking new use cases like capacity, traffic‑matrix, and peering analysis.

NTT

Paolo Lucente

Big Data Architect