Datacor has built a self-service rental analytics system inside TrackAbout, giving gas and welding distributors dashboards and natural-language questions over their own operational data. The useful detail is beneath the AI gloss: cross-cloud ingestion, tenant isolation and validation are what make the feature safe enough to put near revenue decisions.
AWS AI Watch analysis
What happened
Datacor’s TrackAbout software now combines embedded Amazon Quick Sight dashboards with Amazon Q in Quick Sight, AWS’s generative business-intelligence capability. Users can ask questions such as which customers have the most unreturned cylinders, how rental revenue breaks down, or how many assets have been out for more than 90 days, rather than waiting for a custom report from IT.
The system moves data from TrackAbout’s Azure SQL environment into AWS through Azure Data Factory, Amazon S3, AWS Glue, Lambda and Step Functions. Quick Sight SPICE datasets are refreshed after processing, while row-level security limits each dashboard and natural-language query to the relevant customer’s data.
Datacor says the rollout began with a data pipeline in July 2025, reached a pilot of six or seven customers in September, and was released to TrackAbout customers using rental functionality in November 2025. The AWS case study was published on 25 September 2026.
Why it matters
This is a more interesting AI story than another chatbot stapled to a dashboard. The practical gain is that distributors can investigate fleet utilisation, billing exceptions, asset ageing and revenue recovery in the same meeting, without filing an IT ticket for every follow-up question.
The architecture also shows where enterprise AI earns its keep. The natural-language layer is only useful if the data arrives reliably, customer records remain separated and the system can explain enough of the answer to support a commercial decision. Datacor says it tested business terms such as “lost cylinders”, “demurrage” and “fill cycle”, and used pre-built measures for metrics including average rental duration and revenue recovery rate.
Our read
The strongest lesson here is gloriously unglamorous: trustworthy AI often looks like plumbing with a search box attached. Datacor’s own account is a vendor case study, so its claims about accuracy and customer reception should be treated as company-reported rather than independent performance evidence. Still, the design gives readers a useful checklist for judging similar systems: data lineage, tenant-level access controls, validation, refresh behaviour and a clear answer to the question of what happens when the model is wrong.
What to watch
- Whether Datacor publishes independent measures of answer accuracy and revenue impact.
- How quickly the cross-cloud pipeline refreshes data in day-to-day use.
- Whether row-level security is tested against natural-language queries as rigorously as dashboards.
- Whether customers use the what-if analysis for decisions, rather than merely admiring a shinier report. The AI may get the headline, but the access controls are the adult supervision.
Discussion spark: For business AI handling revenue data, which matters more: a smarter natural-language model or stricter data pipelines and access controls?
Sources and evidence
- How Datacor built self-service rental analytics with Amazon Quick Sight (25 September 2026, 15:54 UTC)
- Deploying real-time personalized speech with Qwen3-TTS on Amazon SageMaker AI (25 September 2026, 15:54 UTC)
not affiliated with or endorsed by Amazon Web Services (AWS)