Privacy

ATLAS is operated by Cartooli Business Ventures Inc.. Privacy, access, and deletion requests go through the contact form. We do not publish a privacy email address.

The ATLAS app at /app stores fund configuration and portfolio inputs in your browser's local storage on your device. That data is not sent to ATLAS servers unless you use a feature described below that sends content to our API.

At /app/import, most file types are parsed in your browser. PDF uploads are sent to a server endpoint for text extraction; URL fetches go through a server endpoint that retrieves the page you specify. Extracted text is held in your browser session until you run an analysis.

When you submit an analysis, the composed prompt (your pasted text plus extracted content from your sources) is sent to our /api/chat route, which forwards it to Anthropic's API for processing. For Excel workbooks, structure-only profiling is the default; if you opt in, cell values from those workbooks are included in what is sent to Anthropic.

At /app/advisor, your chat messages and a summary of your current fund settings are sent through /api/chat to Anthropic for a response. Chat history for a session is kept in the page while you use it; we do not describe server-side chat logs in this codebase.

You may optionally supply your own Anthropic API key in the advisor or import settings. That key is sent only inside requests to our API route so Anthropic can be called; it is held in page state for the session, not written to local storage in the code paths we reviewed.

If you do not supply a key, the site may use a platform Anthropic API key when configured. Those requests are rate-limited per IP address on the server.

Messages you submit on the contact form (name, email, subject, and message) are stored on the server in a JSON file in development or in Upstash Redis/KV when those credentials are configured in production. The server also records a timestamp and the client IP address from request headers.

The positioning feedback survey collects your role, optional organization and AUM band, ratings, willingness-to-pay bands, optional free-text answers, and optional name and email if you choose to provide them. Submissions are stored on the server in Upstash Redis/KV or a local JSON file, with a timestamp and client IP. When Resend is configured, a summary may be emailed to operator-configured notification addresses; delivery is best-effort.

The VC/PE survey collects firm type, AUM band, pain level, feature priorities, budget owner, pricing reaction, whether you would pay, optional free-text fields, and optional email. Storage and optional Resend notifications follow the same pattern as positioning feedback.

The fund tools may show a short lead-qualification prompt that posts to /api/lead-qualification. That collects optional role, fund profile, pain areas, current tooling, timeline, optional email, source label, timestamp, and IP, stored in Redis/KV or a local JSON file.

The services builder saves quotes you generate (company and contact details, fund stage, notes, selected service functions, and computed monthly totals) through /api/services/quote into the same Redis/KV or file-backed key-value store. Starting a trial through /api/services/onboarding links your quote to an onboarding record (contact email, trial start, and quoted monthly total). When Stripe is configured and the quoted total is greater than zero, checkout is completed through Stripe; otherwise the trial is recorded without card collection in the application store.

The site is hosted on Vercel. Server-side storage uses Upstash Redis/KV when configured. AI features send prompts to Anthropic as described above. Survey notifications may use Resend. Services checkout may use Stripe when enabled. Vercel Analytics runs on public pages as noted below. We do not sell personal information through these integrations.

We use Vercel Analytics on public pages. It collects usage metrics as described in Vercel's documentation. We do not add extra analytics identifiers in the application code beyond that integration.

Some internal routes use HTTP Basic authentication when environment credentials are set. Access to those areas is limited to configured operators and is not part of the public fund tools.

The public fund tools do not create a login account on our servers. Portfolio and fund settings you enter stay in your browser unless you use a feature that calls our API, as described above. Services onboarding stores business contact and quote records tied to your email for that flow; it is not a general-purpose user profile for the fund app.

Contact messages, survey responses, and lead-qualification records are kept until we remove them. Each list is capped in software (10,000 entries for contact and the two public surveys; 50,000 for lead qualification); when a cap is exceeded, the oldest entries may be dropped automatically. Service quotes and onboarding records persist in the key-value store until deleted as part of operations.

When you use AI features with the platform API key, we apply a per-IP request limit in server memory that resets over time; we do not store your full chat content on the server for that session. Our chat API writes structured audit metadata (model, timing, token counts, and a short preview of the last user message) to application logs captured by the host; it does not log full prompts or responses.

You can clear fund tool data stored in your browser through the app settings or by clearing site data for this domain in your browser.

To request access to or deletion of information you submitted through the contact form, surveys, lead capture, or services flows, use the contact form with enough detail for us to find your submission. We will process verified deletion requests within 30 days where we can reasonably identify your data in our stores.