Solely for your recommendations
Your data is used to tailor Y's analyses and recommendations to your context — your maturity, your priorities, your sector. Nothing else.
Your data stays yours. Encrypted in transit and at rest, client-isolated — and never used to train a model.
For institutions and enterprise accounts, confidentiality is the foundation. AES-256 at rest, TLS 1.3 in transit, client isolation, no training on your data. AI processing relies on the providers identified for each service and may take place outside the EU, framed by Standard Contractual Clauses — see the sub-processor list.
GDPR compliant · DPA · SOC 2 (sub-processors) / ISO 27001 certified infrastructure (AWS, Supabase) · Per-project isolation
No fine print. Here, in plain language, is the exact use of your data at YPrtnrs.
Your data is used to tailor Y's analyses and recommendations to your context — your maturity, your priorities, your sector. Nothing else.
Your data is never used to train, fine-tune, or improve any AI model — ours or a third party's. It remains yours, full stop.
No resale, no commercial sharing. Our limited technical sub-processors are contractually bound and listed in our trust center.
Export, rectification, and deletion on request. Strict per-client isolation: your data never intersects with another organization's.
Four stages, one rule: at every step, your data remains encrypted, isolated, and traceable.
Encrypted in transit (TLS 1.3), stored encrypted (AES-256).
Per-client isolated environment — to adapt recommendations, never to train.
Every recommendation sourced and logged; nothing executes without your sign-off.
Full reversibility — export, rectification or deletion on request.
production of AI diagnostics and the AI Blueprint from your answers and authorised documents.
Y’s replies in the application, through our application host’s AI gateway.
web research, voice synthesis, video, presentations and document conversion, each handled by the sub-processor listed for that operation.
The model provider, model and region applicable to each function are those identified in the contractual processing schedule for the service concerned and in the sub-processor list; there is no single fixed mapping between a function and a provider.
Whatever the provider, your data is not sold, is processed only for the purpose of the service, remains confidential and is not used to train AI models. Each provider’s retention periods are those set by its applicable terms and by the processing schedule.
Persistent storage remains in the European Union. Where AI processing takes place outside the EU/EEA, it happens only within the existing contractual safeguards, including the EU Standard Contractual Clauses (Decision 2021/914), with encryption in transit and data minimisation.
For information, the analysis engine’s code documents an integration with Mistral AI. This describes a technical capability: it does not add a sub-processor and is not a commitment on region or retention.
Everything your legal and security teams need to complete their review — without reaching out to us.
GDPR addendum, ready to sign
Data Processing Agreement (DPA)Technical third parties & locations
Sub-processor listTechnical & organisational measures
Usage inventory, AI charter, training register (art. 4), transparency (art. 50)
Kept up to date under the CAIO mandate →Data usage & individual rights
Privacy policySOC 2 (sub-processors) · ISO 27001, on request
Request attestationsPublisher & hosting provider
Legal noticesThe model provider, model and region applicable to each function are those identified in the contractual processing schedule for the service concerned and in the sub-processor list; there is no single fixed mapping between a function and a provider. Persistent storage remains in the European Union. Where AI processing takes place outside the EU/EEA, it happens only within the existing contractual safeguards, including the EU Standard Contractual Clauses (Decision 2021/914), with encryption in transit and data minimisation.
Your security team needs a dedicated conversation? Let's talk. Otherwise, the simplest place to start is AI Pulse — free, around 10 minutes.
Human-supervised AI · traced sources · every figure traced to its source.