Services
Simverk works with the engineering, reliability and operations teams behind production systems, fleets and infrastructure: building the model, running the analysis, or building the tool that lets your own people do it repeatedly.
The work applies wherever physical systems have to perform — manufacturing, energy, transport, defence and process industries. It suits questions that are technically specific and commercially significant, where the answer depends on how a system actually behaves over time rather than on a single calculation. Projects are deliberately small and scoped.
Simulation modelling
Discrete-event and agent-based models, built to answer a specific question.
- Discrete-event models in .NET and Python
- Agent-based models
- Capacity and throughput analysis
- Bottleneck identification
- Scenario and what-if studies
- Experiment design and results analysis
- Verification and validation
- Review of existing models
- Built on open-source foundations, not a licensed platform
Reliability engineering
Understanding how systems fail, and what to do about it.
- System reliability and availability modelling
- Failure and maintenance data analysis
- Survival and time-to-event analysis
- Maintenance and inspection interval studies
- Spares and support requirements
- Reliability input to design decisions
Engineering analytics
Making engineering and operational data answer the question that was actually asked.
- Exploratory analysis of engineering data
- Statistical modelling and inference
- Machine learning where a simpler model will not do
- Degradation and remaining-life estimation
- Uncertainty quantified rather than hidden
- Reproducible analysis pipelines
Decision support tools
Putting a model or an analysis into the hands of the people who use it.
- Native desktop applications for Windows, Linux and macOS (Avalonia)
- Web applications (Django)
- Interfaces for non-specialist users
- Reporting and dashboards
- Packaging an existing analysis as a tool
Workflow automation and data systems
Removing the manual steps between data, analysis and decision.
- Automation of repetitive analysis pipelines
- Engineering data models and storage
- Integration between simulation, analysis and reporting
- Batch experiment execution
- Migration of spreadsheet-based analysis
LLM-accessible tooling
Exposing models and analyses so an LLM agent can drive them.
- MCP servers over simulation and reliability models
- Command-line interfaces, so a model is scriptable and agent-drivable
- Structured APIs and machine-readable outputs
- Deterministic tools an agent calls, with the engineering logic staying in the model
- Local inference, so no data leaves your network (llama.cpp)
- Cloud models where scale or frontier capability justifies it
How engagements are shaped
Every engagement has a defined end point. These are the shapes it usually takes.
Feasibility study
A short piece of work establishing whether an approach will pay off, before anyone commits to building it.
Focused project
Fixed scope, a defined deliverable and an agreed timeframe. The most common way Simverk works.
Tool build
A working application, handed over with source code and documentation.
Review and advisory
A technical review of an existing model, analysis or approach, with written findings.