Trust, Data Handling and Security

What Layerz does and does not do

Layerz is a financial modeling and calculation workspace. It stores your models, computes them, and serves them to the AI agent you choose to connect (Claude, ChatGPT, Le Chat, Codex or any other MCP client) over MCP or the API. Layerz is not an AI provider: it does not run AI on your data on your behalf.

Your data is not used to train AI

Layerz does not train any AI model on your data, and does not send your model content to AI providers. AI processing only happens when you connect your own AI agent, under your own AI account, governed by your agreement with that AI provider.

AI Act positioning

Because the AI runs under your own account, Layerz is not the provider or the deployer of that AI system. When you connect your own AI agent, you act as its deployer and remain responsible for how you use its outputs. A use case may become high-risk if outputs are used for credit scoring, lending, insurance, employment, access to essential services or other decisions affecting natural persons.

Human oversight

• Every model change, including changes made through a connected AI agent, is recorded in the version history, remains editable and can be undone.
• You must validate assumptions, source data, formulas, outputs and exports.
• Model outputs may contain errors and should be reviewed before use.

Who can see your models

• Each project lives in its own workspace. A model is visible only to the members of its workspace and to the people you share it with.
• Someone you invite on a model sees that model, not your other projects.
• A public link is off by default, and you can switch it off at any time.

Traceability

Layerz keeps each model’s version history and logs API and MCP requests, so you can see what changed, when and through which channel. How far back the history reaches depends on the plan of the workspace that holds the model.

Responsible use boundaries

• Do not use Layerz as the sole basis for decisions affecting individuals.
• Do not use Layerz for social scoring, manipulation or prohibited AI practices.
• If you use outputs in regulated workflows, you must add your own controls, documentation and human review.

Security

Your data is encrypted in transit (TLS) and at rest (AES-256), and stored in the European Union (Ireland). Layerz also uses authentication, hashed API keys, encrypted credentials for connected services, access controls, row-level security, restricted admin access, log redaction and error monitoring.

Contact

For legal, security or data protection questions, contact [email protected].