Data Engineering · 3 min read
Makro consolidates 10+ data tools onto Tencent Cloud, claims 25% efficiency gain
Makro and Tencent Cloud report completing Southeast Asia's first data platform consolidation at this scale: 10+ tools onto two platforms, 25% efficiency gain, 10,000 daily tasks — all vendor-verified.
Key takeaways
- Makro consolidated 10+ data products onto Tencent Cloud's Data Lake Compute and WeData platforms.
- Nearly 10,000 production tasks now run daily on the consolidated platform.
- Vendor reports a 25%+ rise in downstream operational efficiency across 20+ business units.
- SQL table validation time fell from 1–2 hours to about 5 minutes using Tencent's CodeBuddy.
- Tencent Cloud calls it the first consolidation of this scale in Southeast Asia.

Thailand's Makro has finished consolidating a data estate of more than 10 vendor products onto two Tencent Cloud platforms, with the vendor claiming zero data discrepancies and no business interruption during migration. Tencent Cloud stated on Tuesday that this is the first completion of such scale in Southeast Asia. Both claims come from the vendor's own announcement; no third-party audit is cited.
What was actually consolidated?
Makro, the wholesale arm of CP AXTRA and part of CP Group, ran data infrastructure scattered across more than 10 products from multiple vendors. Transaction, product and membership data grew alongside the company's digital business and its ASEAN expansion, and the legacy setup needed modernising.
The consolidation landed on two platforms:
- Data Lake Compute — Tencent Cloud's serverless data lake analytics service.
- WeData — its platform for data integration, development, orchestration and governance.
Makro now runs data engineering, analytics and AI workflows on a single architecture. The same data serves both daily operations and AI model development. Planned use cases include warehouse forecasting and inventory optimisation.
What do the numbers say?
Tencent Cloud reports measurable outcomes, though all figures are self-reported:
- Downstream operational efficiency up more than 25 percent.
- More than 20 business units, including supply chain and finance, completed data modernisation.
- Nearly 10,000 production tasks run daily on the big data platform.
- Data governance was standardised, with tables, fields and relationships centrally managed and traceable to their origins.
A 25 percent efficiency gain across 20 business units is a strong claim. The statement does not define how "operational efficiency" was measured, over what baseline, or by whom. Treat the figure as directional until Makro publishes methodology.
Where did the tooling come from?
Tencent Cloud used CodeBuddy, its AI-assisted coding tool, to analyse business SQL scripts, convert code and validate data. The vendor reports that validating one data table, previously a one-to-two hour task, now takes about five minutes. That is roughly a 95 percent reduction in validation time per table — again, a vendor-reported benchmark with no stated sample size or error rate.
What did the executives say?
Tanit Chearavanont, group chief wholesale business officer of CP AXTRA, said Tencent Cloud demonstrated strong technical expertise and a clear understanding of Makro's business needs.
Poshu Yeung, senior vice president of Tencent Cloud and head of Tencent Cloud International, said large-scale data modernisation is challenging, but the project shows enterprises can consolidate complex environments while remaining stable. Note what Yeung did not say: no timeline, no migration cost, and no downtime figures appear anywhere in the statement.
What should data teams take from this?
The case is a useful reference point for regional retailers sitting on fragmented, multi-vendor data stacks. A single architecture serving both operational analytics and AI development removes the twin-pipeline problem that inflates cost and slows model iteration. Standardised, traceable governance — centrally managed tables, fields and relationships — is the prerequisite for that, not an afterthought.
The caveats are equally instructive. The announcement omits project duration, budget, and any independent verification of the zero-discrepancy claim. Vendors rarely publish failed migrations in the same market; survivorship bias applies. Data teams evaluating similar consolidations should demand baseline definitions, measurement windows and rollback evidence before treating this as a template.
Google News: Data Engineering (Source)
More from Tom Whitfield
Show full bio
Staff writer covering regulation, policy and their impact on Analytics Digest is a professional daily at Analytics Digest.
16 articles