Privacy Policy

This policy explains what personal data ScaleLogix AI handles, why we handle it, who we share it with, and what choices you have. It covers the ScaleLogix AI OS platform — the marketing site, the AI Revenue Audit funnel, the agency dashboard, the client portal, and the AI voice, chat, website, and outreach services delivered through them.

The platform is sold to agencies, which in turn serve their own business clients, which in turn serve consumers. Because a single record can belong to any of those layers, section 2 sets out who is responsible for what before the rest of the policy describes the data itself.

ScaleLogix AI OS is multi-tenant. Data sits at one of four levels, and the responsible party differs at each:

In plain terms: for your own ScaleLogix account we decide how data is used. For the customer and consumer data you load into or generate through the platform, youdecide, and we act on your instructions. If you are a consumer who spoke to an AI voice agent or filled in a form on a business's website, the business you contacted is your first point of contact — see section 12.

Account and organisation data

Authentication is handled by Supabase Auth, which stores your email address, your name, and your credentials. Alongside it we store your organisation record (billing and contact email, contact phone), your role and workspace memberships, and pending invitations.

Access requests

The platform is invite-only. When you ask for access we collect your name, email address, phone number, company, your message, your answers to the qualifying questions, and — for abuse prevention on a public, unauthenticated form — your IP address and browser user agent.

AI Revenue Audit

The audit funnel collects business-profile answers (industry, business model, team size, lead sources, current CRM, and the obstacles you describe in your own words), your business name, and an email address if you ask for the results. It also stores the AI-generated report produced from those answers.

Customer and consumer data you bring

Operating the platform means storing data about your clients and their customers on your behalf:

• Client records — company name, contact name, email, phone, industry, website, location, and your own notes.
• Calls — caller name, email, and phone number; the full transcript; an AI summary; a sentiment rating; a link to the recording; and any structured fields the agent extracted from the conversation. See section 4.
• Appointments — contact name, email, phone, and the booked time.
• Leads — name, email, phone, company, industry, location, the source channel, and the complete original payload received from the lead source.
• Website chat— the visitor's name where given, and the full text of the conversation, which may contain whatever the visitor chooses to type.

Technical and security data

We record IP address, user agent, and referrer when an embedded analytics dashboard is opened, so that access to a client's reporting can be audited after the fact. We use IP addresses transiently to rate-limit requests. Where you give marketing consent we store the IP address and user agent as proof of that consent, and we keep a delivery ledger of transactional emails that records the subject line but never the message body.

Calls handled by a ScaleLogix AI voice agent are recorded and transcribed. The voice service is provided by Retell AI, which captures the audio, produces the transcript, and analyses the conversation. After a call ends, Retell sends us the transcript, an AI-generated summary, a sentiment rating, a link to the recording it hosts, and any details the agent captured — typically the caller's name, email address, phone number, and whether an appointment was booked. We store those on the relevant workspace record. The audio itself remains hosted by Retell; we store the link to it.

Retell supports masking personal details in stored records, and the platform default is post-call redaction, applied by Retell before the record is retained. An operator can change that setting per agent, and turning it off means raw, unmasked transcripts are retained.

GoHighLevel is the system of record for contact and appointment data. The agency connects its own GoHighLevel account, and the platform writes into it rather than replacing it. Specifically:

• Contact details captured on a call or in chat — name, email address, phone number — are created or updated as GoHighLevel contacts.
• Call outcomes are written onto the GoHighLevel contact record as custom fields: booking status, sentiment, recording URL, call summary, and the full transcript. The platform creates those fields automatically if they do not already exist.
• Appointment bookings made by an AI agent send the contact's full name, email address, phone number, preferred time, and time zone into the client's GoHighLevel calendar.

Because the GoHighLevel account belongs to the agency, data written there is governed by the agency's own agreement with GoHighLevel and by the agency's own retention settings — not by ours. The same applies to the separate analytics database provisioned in the agency's own infrastructure, which holds a second copy of call, appointment, and chat records.

Several features send content to large language models operated by third parties.

• AI assistant— the in-app and Slack assistant runs on Anthropic's Claude models. It can use tools that read contacts, conversations, calendars, and call records, which means the personal data those tools return becomes part of the prompt sent to Anthropic.
• Website builder — your business description, brand details, uploaded imagery, and the text of your conversation with the builder are sent to Claude to generate and revise the site.
• AI Revenue Audit — your intake answers are sent to Claude to produce the report.
• Document summarisation — text extracted from documents you upload for summarisation is sent to Claude.
• Knowledge base search — knowledge-base content is converted into embeddings by OpenAI and stored as vectors in Pinecone, using API keys the agency supplies from its own accounts.

Anthropic is accessed under ScaleLogix's own account. OpenAI and Pinecone are accessed under the agency's account, using credentials the agency provides.

We use the providers below to run the platform. The final column matters: where the account holder is the customer, you have connected your own account, you hold the relationship with that provider, and their terms — not ours — govern what happens to the data there.

Two clarifications, because both are commonly assumed and neither is accurate. There is no Google Calendar integration — all calendar functionality runs through GoHighLevel. The Instantly and HeyReach integrations are read-and-control only: the platform reads campaign statistics and can start or pause a campaign, but it does not push prospect lists into either service.

We use a small number of first-party cookies and browser storage keys, all of them functional. We do not use advertising cookies, and there is no Google Analytics, no advertising pixel, and no cross-site tracking on this platform.

• Session cookies — set by Supabase Auth to keep you signed in. Strictly necessary.
• Branding cookie— a cookie holding your user identifier for 24 hours so your agency's branding renders without a flash of the wrong theme on first paint.
• Local and session storage — your theme preference, cached branding, sidebar and navigation state, and drafts of forms you have not yet submitted, so that work is not lost on a refresh. This data stays in your browser.

Page-view and performance measurement uses Vercel Web Analytics, which is cookie-less and does not build a profile of you across sites.

We use Sentry to detect and diagnose faults. When an unhandled error occurs, Sentry receives the error, a stack trace, and context about what the application was doing. Sentry is not configured to attach IP addresses or cookies to those reports.

In production, when a session encounters an error, Sentry also captures a replay of that session — a reconstruction of what was on screen around the failure. Ordinary sessions are never recorded; only sessions that hit an error are.

Access to customer data is enforced on the server by role, not merely hidden in the interface, and the database applies row-level security so one tenant cannot read another's records. Third-party credentials you store with us — API keys and tokens for your own CRM, voice, and automation accounts — are encrypted at rest with AES-256-GCM. Sensitive actions such as deployments, credential changes, and dashboard access are recorded in an audit trail. Requests are rate-limited, and production and non-production environments are kept separate.

No system is perfectly secure. You are responsible for keeping your own credentials safe and for removing access promptly when a team member leaves.

We keep account and organisation data for as long as your account is open. Operational records — calls, transcripts, appointments, leads, chat conversations, and activity history — are retained so the platform can show you your own history, and are deleted when the workspace they belong to is deleted, which removes the associated records along with it. Billing history is retained separately, because we are required to keep records of transactions.

Depending on where you live, you may have the right to access a copy of your personal data, to correct it, to delete it, to receive it in a portable format, to object to or restrict certain processing, and to withdraw consent you have given. You will never be treated differently for exercising these rights.

If you are a consumer

If you spoke to an AI voice agent, used a website chat, or submitted a form on a business's website, that business decides how your data is used, and we act on its instructions. Contact that business first. If you contact us instead, we will pass your request to the relevant customer and support them in answering it.

Marketing email

Marketing email requires a confirmed opt-in: ticking a box is treated only as an intent until you confirm from the address itself. Every marketing message carries a one-click unsubscribe, and unsubscribing suppresses that address. Transactional messages — invitations, password resets, and service notifications — are not marketing and are sent as part of operating your account.

Our providers operate internationally, so your data may be processed outside the country where you are located.

ScaleLogix AI OS is a business tool. It is not directed at children and we do not knowingly collect personal data from them. If you believe a child has provided us with personal data, contact us and we will delete it.

We will update this policy as the product changes. Every version carries a dated identifier, shown at the top and bottom of this page, which is recorded alongside your acceptance so it is always possible to establish which wording applied at a given time. Where a change materially affects your rights, we will give notice before it takes effect.

If you have a question about this policy, or wish to exercise a right described in section 12, get in touch and we will respond.