MineLogX AI

How it works

From scattered data to answers you can check.

What MineLogX connects to today, how your data is handled, and answers to the questions industrial and IT teams ask first.

Step 1

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.

Step 2

Validate and organize

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.

Step 3

Ask and act

Dashboards, KPIs, and anomaly detection show what needs attention. Plain-language questions return answers that cite the record or document behind them.

Integrations

What connects today, and what is next.

IntegrationStatus
MQTT (AWS IoT Core)Live telemetryLive
Apache KafkaLive telemetryLive
Batch files: CSV, JSON, NDJSONBulk and historical loadsLive
REST ingestion APIPush from your systemsIn development
PDF documentsManuals, procedures, regulationsLive
OPC UA, ModbusPLCs, SCADA, metersThrough a gateway
OSI PI, AspenTech IP21HistoriansThrough a gateway
IBM MaximoAsset and work managementIn development
Power BIReportingOn request
SnowflakeData sharingRoadmap
Microsoft Azure deploymentHostingRoadmap

Integration scope is confirmed for each customer. Third-party names identify compatible technologies and do not imply endorsement or partnership.

How your data is handled.

  • Each customer's data is kept separate, down to the row level in the curated data layer.
  • Device connections use per-tenant certificates; a device can only publish to its own customer's topics.
  • Every AI call passes through guardrails that block prompt-injection attempts and filter personal information.
  • Answers cite the record or document they came from, and say when the data is not there.
  • AWS is the reference deployment. Customer-managed and local-model deployments are scoped per project.

FAQ

Common questions

What data sources can MineLogX connect to?

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.

Where does my data live?

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.

How do the AI answers stay grounded?

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.

Does MineLogX replace IBM Maximo?

No. Maximo stays your asset and work-management system of record. MineLogX adds operational context, such as condition signals and maintenance history, around it.

Which AI models does it use?

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.

How do we get started?

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.

Have a question that isn't here?

Book a demo, or write to demo@minelogxai.com.

Book a demo