Data and security
What happens to your data when we automate your process
An automation touches your CRM, your inbox, your customers. These are the commitments every Technosys engagement runs under, written in plain language so your IT and legal teams can read them in five minutes.
Your data is never used to train a model
We call commercial AI models through APIs with data-retention and no-training terms. Nothing you give us becomes part of a model, ours or anyone else's.
A person approves customer-facing output
Emails, quotes, replies and anything a customer will read are drafted by the workflow and approved by a human, unless you decide otherwise for a specific low-risk case. Billing, refunds, complaints, security and health matters always escalate to a person.
Least-privilege access
Integrations use scoped API keys or OAuth with the minimum permissions the workflow needs. Keys live in the automation platform's credential store, never in code, documents or chat.
Compliant hosting where it matters
Healthcare, finance and other regulated work runs on infrastructure that meets the relevant requirements, with signed agreements, encryption in transit and at rest, and audit logging. Protected health information is never sent to a general-purpose AI model.
You can see everything
Every workflow ships with a run log and a monitoring dashboard. You can see what ran, what it decided, what a person approved and what failed.
Minimal retention, clean exit
We keep only the data a workflow needs, for as long as it needs it. When an engagement ends, credentials are revoked, copies are deleted and the workflows are handed over with documentation.
Outbound that protects your domain
Lead generation runs from dedicated sending domains with warm-up, bounce control and daily caps. Your primary domain is never used for cold email.
Your systems stay yours
Workflows are built inside your Make.com, n8n, HubSpot or Salesforce accounts wherever possible, so nothing is locked to us.
Questions IT and legal teams ask
Which AI providers do you use?
OpenAI and Anthropic (Claude) through their APIs, under business terms that exclude training on customer data. For some clients we run open models on private infrastructure when data cannot leave their environment.
Will you sign an NDA or data processing agreement?
Yes. We sign NDAs before an audit and a data processing agreement before any production work that touches personal data.
Who at Technosys can see our data?
Only the engineers assigned to your engagement, under role-based access that is revoked when they roll off. Access is logged.
Can we host the automation ourselves?
Yes. Self-hosted n8n and private model deployments are common for regulated clients. We build, you host, and we support it.
Need a specific control documented for a vendor review? Email info@technosysmanagement.com and we will send the relevant section.
Ready to talk through your process?
The audit starts with which systems are involved and what data moves between them. Security questions are welcome on the first call.