Product vision
Weissr starts with where the company needs to be in ten years, and what capital decisions get it there. AI agents will run tomorrow’s operating systems within constraints that remain a management decision. Weissr Capex Strategy is the conscious brain.
The organism
It is the same split nature already made between the autonomic nervous system and the cortex.
Level 1: the conscious brain
Where the company goes, which assets it holds, where capital is placed.
These choices define scope (what the business does) and time horizon (which assets it still owns ten years from now). They remain with management because they define the company’s long-term direction.
In capital-intensive companies, that framework is the capex strategy. Weissr Capex Strategy is therefore the conscious brain, and the enabler of every autonomous system below it.
Level 2: decision support
Agents generate options, model scenarios, surface risks and run sensitivities.
Even the brain can be heavily AI-assisted. The final decision is different. These are long-cycle, irreversible decisions with strategic consequences that unfold over years. The framework the autonomous body operates within is set by people, deliberately, a few times a year.
Level 3: the autonomous body
Supply chain, finance, procurement: run end-to-end by AI agents within a given framework.
Scale does not equal command
“It’s about realising that the capex strategy is the enterprise’s strategy.”
John Williams, CEO, Paper Excellence
ERP, supply chain, asset management and procurement are larger than capex by every easy measure, and still downstream of it. They process the present and optimise within the frame. They cannot set the strategy.
Downstream systems
Most capex software focuses here: project tracking, budget control and ERP integration. These systems operate within the direction set by Capex Strategy.
Strategic direction
The decision stays human
AI can prepare the brain’s work in striking detail. It cannot take its place. Four reasons the framework requires a human signature.
A €500m plant plays out over decades. AI learns by iterating on feedback. Here there is none. Get it wrong and you do not iterate. You live with it. That asymmetry of error is why the decision demands a human signature.
Returns vs decarbonisation. Local jobs vs offshoring. Community impact vs shareholder yield. These are not optimisations. They are choices about which future the company is willing to fund. Choosing among incommensurate values is the working definition of judgement.
Three new mills shift pulp prices. A regional decarbonisation programme moves the regulatory landscape. AI models are trained on historical data: they optimise within a world assumed to stay roughly the same. Capex strategy assumes that world will shift in response to your own bets.
When €100m+ ends up in a stranded asset, a human answers to the board, the regulator and the capital markets. You cannot hold an agent accountable. The EU AI Act already classes capital-decision AI as high-risk.
How it works
At the centre sits the Asset Base: what the company owns, operates and is investing in, including the future portfolio encoded in its capital decisions. Three connected stages continuously update it.
Asset Base: the shared system of record. The learning loop closes here.
Operating model
Five agents, each with a clearly scoped role and a defined interface to the Asset Base. Users delegate intent through the Decision Support Agent.
Generate, challenge and compare strategy options.
Build and update the capital budget under the active strategy.
Approve, execute and track projects. Updates actuals.
The conversational entry point for the human user.
Import and export data via MCP secure connectors.
What the user can ask
In developmentIntegration
The Decision Hub connects to your system landscape through MCP (Model Context Protocol) secure connectors, an open standard for connecting AI agents to external systems. Data flows in both directions.
Master data, governance, productivity
The physical asset perimeter
A second MCP connector exposes the hub to the user, through a conversational interface for answers, recommendations and approval requests. Customer-side or third-party agents can operate against the same surface.
See current integrationsWhat this means
01
The place where the strategic framework is set.
The product becomes the place where the strategic framework is set and continuously updated. Not a passive book of investments.
02
Every action and data object exposed as an MCP tool.
Every workflow action and data object is exposed as an MCP tool, so customer-side and third-party agents can operate against the hub natively.
03
The agent fleet is the product surface.
The UI shrinks to a thin admin and visualisation layer for governance and exception handling.
04
Strategy, budget and execution on one model.
Strategy, budget and execution all converge on the same shared model, closing the learning loop between plan and outcome.
05
Keep the systems you already use.
The hub connects to ERP, project management, BI, Office, supply chain, CAD, asset management and procurement through MCP, so customers can retain the systems they already use.
Use thirty minutes to examine where capital is going and whether it matches the company’s long-term direction.
Book a demo