A Brazilian trucking company — a fleet of ~150 trucks, running over 10,000 trips and issuing more than 20,000 CT-e freight documents per month — operated with data scattered across systems that didn't talk to each other: a TMS for freight and shipping manifests, fleet telematics for tracking, and an ERP for finance. The consolidated view of the operation lived in hand-built spreadsheets, up to 7 days behind reality.
Client anonymized by request.
The challenge
- Reconciling telematics with TMS records was a recurring manual nightmare: trucks constantly changed device IDs in the tracking platform, and trip start/end dates rarely matched across systems.
- Cost per kilometer, fleet utilization and delivery SLA were calculated by hand; management received spreadsheet reports with up to a 7-day lag.
- Analysts spent precious hours every week just assembling pivot tables — not analyzing the operation.
The solution
Meta Dados integrated the sources and delivered self-hosted Metabase as the single visualization layer:
- Integration — automated pipelines connecting TMS, telematics and ERP into a central analytical database, with vehicle ID mapping and trip reconciliation resolved at load time.
- Modeling — a layer unifying trip, freight, vehicle and driver, with the "trip" as the central grain.
- Governance — on open-source Metabase, each branch is an isolated connection (its own schema in the data warehouse), with groups and collection-level permissions mapping who sees what: each unit manager sees strictly their own operation, while executives track the entire fleet through consolidated executive dashboards.
- Dashboards — operations (trips in progress, utilization, on-time performance) and finance (cost per km, margin by route and customer, freight reconciliation).
Results
Decision-making time dropped drastically: KPIs like cost per kilometer, fleet capacity utilization and delivery punctuality moved from spreadsheets that ran days behind to daily monitoring. The handover was complete — today the carrier's internal team builds its own Metabase questions independently, backed by Meta Dados' ongoing managed support and platform evolution.
The stack
TMS · fleet telematics · PostgreSQL · self-hosted Metabase · Python.
Why Metabase
Self-hosted and open source — sensitive operational data stays on the client's infrastructure; isolated connections and group-based permissions to segregate business units; and a learning curve that let the carrier's own team create their questions after the handover.
Frequently asked questions
Do you need to replace the TMS or the telematics platform to get BI?
No. The work is integrating the sources the operation already has — TMS, telematics and ERP — into a single analytical layer. The source systems stay exactly as they are.
Why cross-reference telematics with the TMS?
Each system knows one part of the story. Telematics knows where the truck actually was; the TMS knows what was planned and billed. Crossed together, they answer what each route truly cost and where the operation loses margin.