Dash Media Limited

Trust Center

FLAME is operated by Dash Media Limited. Where your data lives, who can see it, which services we rely on, and how we protect it.

Last updated
27 October 2026

Overview

FLAME is a workplace performance program for leaders and teams. The FLAME app is the measurement companion to the program: participants complete a short assessment, and the app produces individual and team reports.

The app is small by design. It holds the minimum data needed to produce a report, runs on established cloud infrastructure in Australia, and is maintained by a single named operator. We don't hold payment data, we don't sell or share participant data, and we delete it when the engagement ends.

Operator: Dash Media Limited, a New Zealand company (NZBN 9429047432260). Registered office: 242a Oceanbeach Road, Mount Maunganui, Bay of Plenty 3116, New Zealand.

Security contact: hello@flame.live. Response within one business day.

Hosting and data location

Layer Provider Region
Application and API Vercel Sydney, Australia (syd1)
Database and authentication Supabase Sydney, Australia (AWS ap-southeast-2)
Backups Supabase Same region as the database
Source code GitHub United States

Participant data is stored and processed in Australia. Static assets (fonts, images, scripts) are served from Vercel's global CDN; no participant data passes through the CDN. The operator works from New Zealand.

What data we hold

Data Why Who can see it
Name and work email address To create your account, send your sign-in code, and label your report You; your team's designated administrator; the operator
Assessment responses (25 factors, plus the optional Operating Style and Pressure check) To produce your individual report and the aggregated team report You; the operator. The person running a team program sees who has joined and checked in, and a team report of averages once at least four people have checked in; never an individual's responses unless that person switches sharing on. A person coached one to one shares results with their coach from the start, is told so when they join, and can switch it off
Organization and team name To group participants into a team report You; your team administrator; the operator
Sign-in timestamps and basic request logs Security and troubleshooting The operator

We don't collect: passwords (sign-in is by one-time code), payment details, dates of birth, health information, government identifiers, or device identifiers beyond standard server logs.

The assessment measures workplace performance factors such as energy, focus, and sense of purpose. It is a development tool, not a clinical instrument. An individual's results reach their employer or coach only if that person switches sharing on; the join screen says so before they enter an email address, and the choice can be reversed at any time from the app's Sharing page.

Authentication and access

Participants sign in with a one-time code sent to their work email address. No password is created or stored. Sessions are managed by Supabase Auth: the access token is renewed every hour over TLS, and a session lasts until the person signs out.

Administrative access to Vercel, Supabase, GitHub, and the operator's email is limited to one person (the operator) and protected by multi-factor authentication on every account. There are no shared accounts. Database access is through the Supabase dashboard or authenticated API only; there is no public database endpoint.

Inside the app, Row Level Security policies in the database restrict every participant to their own records, and team administrators to their own team's aggregated data.

Subprocessors

See the full table at /trust/subprocessors. In short: Vercel (hosting), Supabase (database and authentication), GitHub (source code), Anthropic (AI-generated report narrative), and Twilio SendGrid (the sign-in email). Each publishes its own compliance documentation, linked from the table.

What we send to the AI provider: assessment scores and the factor labels needed to write a readable narrative. We don't send names, email addresses, or organization names. Anthropic's commercial terms prohibit training on this data, and inputs are deleted from their systems within 30 days.

A subprocessor is a third party that stores or processes participant data on our behalf. This list is current as of 27 October 2026. We update it before adding a new subprocessor and keep a dated change log at the bottom.

Provider Service What they process Region Compliance
Vercel Inc. Application hosting, serverless functions, CDN All application traffic, including participant data in transit through the application layer; request logs Functions: Sydney, Australia (syd1). CDN: global, static assets only SOC 2 Type II, ISO 27001. security.vercel.com
Supabase Inc. PostgreSQL database, authentication, backups All stored participant data; email addresses for one-time code delivery; backups Sydney, Australia (AWS ap-southeast-2) SOC 2 Type II, HIPAA (available). supabase.com/security
GitHub Inc. (Microsoft) Source code hosting, CI Source code only. No participant data United States SOC 2 Type II, ISO 27001. github.com/security
Anthropic PBC Claude API, used to write the narrative text in reports Assessment scores and factor labels. No names, email addresses, or organization names United States (no regional inference option is set) SOC 2 Type II, ISO 27001, ISO 42001. No training on API data; inputs deleted within 30 days. trust.anthropic.com
Twilio Inc. (SendGrid) Transactional email: the sign-in email, sent through Supabase Auth Email address and the one-time sign-in code United States SOC 2 Type II, ISO 27001. twilio.com/trust-center

Infrastructure beneath the subprocessors

Supabase runs on Amazon Web Services (ap-southeast-2). Vercel runs on AWS for serverless functions. Both providers' compliance documentation covers their use of AWS.

What we don't use

No advertising networks, no data brokers, no session-recording tools, no error-tracking service, no third-party cookies, and no analytics on the app. The public site flame.live uses Vercel Web Analytics (cookieless page counts, no personal data), covered by the Vercel entry above.

Change log

  • 27 October 2026: Initial publication. Vercel function region moved from Washington, D.C. (us-east-1) to Sydney (syd1).

The same table, on its own page: /trust/subprocessors

Data retention and deletion

  • Participant data is kept for the duration of the client engagement. Deletion is on request: a client can ask for their team's data to be deleted at any time, and we action it within 30 days.
  • Clients can request deletion of their team's data at any time. We action requests within 30 days and confirm in writing.
  • Participants can request a copy of their data or its deletion by emailing the security contact. We verify the request against the account email before acting.
  • Backups are retained for 7 days and roll off automatically, so deleted data leaves backups within that window.
  • Aggregated, de-identified statistics (no names, emails, or team identifiers) may be retained for research and benchmarking.

Security practices

  • Encryption in transit (TLS 1.2 or higher, HSTS enforced) and at rest (AES-256 on Supabase and its backups).
  • Multi-factor authentication on every administrative account.
  • Row Level Security on every table holding participant data.
  • Secrets stored in Vercel and Supabase environment configuration, never in source code.
  • All changes go through GitHub with a preview deployment reviewed before production.
  • Dependencies monitored by GitHub Dependabot; security patches applied promptly.
  • Daily automated database backups in the same region.
  • Logs don't contain assessment responses or sign-in codes.
  • Policies reviewed at least annually and after any significant change to the app.

What we don't have, so you don't have to ask: an in-house SOC 2 or ISO 27001 certification. We rely on the certified infrastructure of our subprocessors (linked above) and keep our own footprint small enough to describe on one page.

Policies

  1. 1. Information Security Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF
  2. 2. Access Control and Acceptable Use Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF
  3. 3. Data Protection, Retention and Deletion Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF
  4. 4. Secure Development and Change Management Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF
  5. 5. Incident Response Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF
  6. 6. Business Continuity and Disaster Recovery Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF
  7. 7. Vendor and Subprocessor Management Policy

    Version 1.0 · Effective 27 October 2026

    Download PDF

Each page has a PDF download. For what the app itself tells people, see the privacy notice shown in the FLAME app.

Report a security concern

Email hello@flame.live. Include what you saw, where, and when. We acknowledge within one business day and keep you informed until it's resolved. We don't take legal action against good-faith researchers who report responsibly.