Software and AI for energy and industry
We apply AI in energy, utilities and manufacturing to the work around the physical assets: connection requests, customer service, quoting, quality and maintenance planning. We build the data pipelines, integrations and agents with OT security designed in, and engineers and operators keep control of the installations.
Industries in energy and industry
The work, and where software fits.
Energy companies and manufacturers run expensive, long-lived assets under strict safety and security rules. The operational technology that controls grids, plants and production lines is rightly kept separate, which means the data it produces rarely reaches the people planning maintenance, quoting orders or answering customers.
Software fits in the layer above the control systems. Read-only pipelines bring historian and machine data together with ERP, SAP IS-U and asset management records, and agents and internal tools use that data to prepare work for planners, engineers and service teams. Security reviews, NIS2 duties and IEC 62443 principles shape every design from the start.
How we improve energy and industry.
OT-safe data pipelines
Historian, SCADA and machine data read through segmented, one-way connections and combined with ERP and asset records.
Customer and connection requests
Intake, completeness checks and status updates for grid connections, supplier customer service and outage information.
Quoting and engineering documents
RFQs, drawings, specifications and subsidy paperwork read, structured and prepared for the engineer or office team.
Maintenance and quality knowledge
Manuals, work order history, inspection results and quality reports searchable for technicians and engineers, with sources.
ERP, SAP IS-U and MES integration
Monitored connections between core systems, market processes and the shop floor, with failed messages in a work queue.
Further reading.
Human-in-the-loop AI agents: design patterns for production
The patterns we use to run AI agents safely in production: approval checkpoints, confidence thresholds, tool permissions, audit trails and fallbacks.
4 min read
Embeddings and semantic search: a practical guide for product teams
How embeddings power semantic search, how to choose a model and a vector store, and how to measure whether your results actually improved.
5 min read
LLM evals: how to test AI features before every release
A practical approach to LLM evals: build a test set from real cases, combine code checks with model grading, and block releases that regress.
5 min read
Common questions.
Only to read data, and only through interfaces your OT team approves. Nothing we build writes to SCADA, PLCs or safety systems, and control stays with your operators.
Energy, drinking water and several manufacturing sectors are covered by NIS2, in the Netherlands through the Cyberbeveiligingswet. Your legal team confirms scope; we design access control, logging and supplier security to support your obligations.
Yes. We deploy in your cloud account, an EU region or on premises where OT rules require it, and can use EU-hosted or self-hosted models.
With a document-heavy process close to customers or orders, such as connection requests or RFQ intake. Shop floor and asset data projects often follow, once security and access arrangements are in place.