How Central Group built real-time retail analytics on QuestDB
Central Group streams analytical, e-commerce and marketing data through Kafka into QuestDB to power real-time Grafana dashboards and alerts for its business units, with no added CPU or RAM since going to production.
Central Group is one of Asia's largest retail groups. Millions of people use its properties every day, from many different devices. To run all of it, the group needs a live view of what is happening across its businesses, and that view has to be fast, accurate and stable.
That view is the Business Control Tower: an analytics platform that brings together online and offline data and turns it into dashboards for each business unit. QuestDB is the time-series database at its core.
One live view
Online and offline data in real-time dashboards for every business unit
No hardware creep
No added CPU or RAM since going to production
Familiar tools
Grafana dashboards over standard SQL
Challenge
The Business Control Tower takes in data from Central Group's analytical, e-commerce and marketing APIs. With traffic from millions of users, the database behind it had to absorb a high input rate, answer dashboard queries quickly and recover cleanly from failures.
The team evaluated several systems before choosing QuestDB:
InfluxDB could not accommodate the input scale our APIs demand. Similarly, our in-out requirement was higher than the capacity provided by Google Suite. SingleStore did not provide any sensible recovery mechanisms when the system crashed, and it was also costly. In the end, QuestDB stood out.
Solution
Data flows from the source APIs through change data capture into Kafka, and from Kafka into QuestDB. Grafana connects to QuestDB over the PostgreSQL wire protocol for dashboards, and the same data drives real-time analytics and alerts.
QuestDB's time-series features, including deduplication and its handling of out-of-order and high-cardinality data, let Central Group monitor its growing data in real time without adding hardware.
Grafana integration
First-party integration for real-time dashboards
SQL in real time
Standard SQL with time-series extensions
Cost effective
High throughput on minimal hardware
| Metric | Value |
|---|---|
| Daily users | Millions |
| Data sources | Analytics, e-commerce, marketing |
| Pipeline | CDC and Kafka into QuestDB |
| Hardware added | None |

Results
| Before | After |
|---|---|
| InfluxDB could not keep up | Fastest in the evaluation |
| SingleStore: weak recovery, costly | No added CPU or RAM |