Connect your data
Telemetry arrives over MQTT, Kafka, or REST, or as batch files (CSV, JSON). PDF documents, including scanned pages and tables, are extracted and indexed.
How it works
What MineLogX connects to today, how your data is handled, and answers to the questions industrial and IT teams ask first.
Telemetry arrives over MQTT, Kafka, or REST, or as batch files (CSV, JSON). PDF documents, including scanned pages and tables, are extracted and indexed.
Every record is checked against a data contract. Invalid records are set aside with the reason, never silently dropped. Clean data moves into raw, cleaned, and query-ready layers.
Dashboards, KPIs, and anomaly detection show what needs attention. Plain-language questions return answers that cite the record or document behind them.
Integrations
| Integration | Used for | Status |
|---|---|---|
| MQTT (AWS IoT Core)Live telemetry | Live telemetry | Live |
| Apache KafkaLive telemetry | Live telemetry | Live |
| Batch files: CSV, JSON, NDJSONBulk and historical loads | Bulk and historical loads | Live |
| REST ingestion APIPush from your systems | Push from your systems | In development |
| PDF documentsManuals, procedures, regulations | Manuals, procedures, regulations | Live |
| OPC UA, ModbusPLCs, SCADA, meters | PLCs, SCADA, meters | Through a gateway |
| OSI PI, AspenTech IP21Historians | Historians | Through a gateway |
| IBM MaximoAsset and work management | Asset and work management | In development |
| Power BIReporting | Reporting | On request |
| SnowflakeData sharing | Data sharing | Roadmap |
| Microsoft Azure deploymentHosting | Hosting | Roadmap |
Integration scope is confirmed for each customer. Third-party names identify compatible technologies and do not imply endorsement or partnership.
FAQ
Live today: MQTT, Kafka, batch files, and PDF documents, with a REST API in development. OPC UA, Modbus, OSI PI, and IP21 connect through a gateway at your site. The integration table above shows the current status of each.
In the reference deployment, in your dedicated tenant on AWS, kept separate from every other customer. Customer-managed and local-model options are scoped per project.
Answers are built from your own records and documents and cite their sources. Guardrails run on every AI call, and the AI never reads raw, unvalidated data.
No. Maximo stays your asset and work-management system of record. MineLogX adds operational context, such as condition signals and maintenance history, around it.
Approved models are selected per deployment, including Anthropic Claude, Amazon Nova, and DeepSeek through Amazon Bedrock. Model choice depends on your policy and data boundaries.
Book a 45-minute demo built around one fleet, site, or asset workflow. We use it to agree on the data sources and scope for a next step.
Book a demo, or write to demo@minelogxai.com.