Privacy Notice
coachparin is an automated AI coaching service operated by Nuggt LLP. This notice explains what happens to your information when you try a conversation, create an account or purchase call time.
1. Who is responsible
Nuggt LLP (UEN T25LL0169A) operates the service. Our registered address is 39 Woodlands Close #08-61D, MEGA@WOODLANDS, Singapore 737856. Contact our Data Protection Contact for privacy requests at support@nuggt.io. The same address handles support and legal notices.
coachparin EXECUTIVE COACHING is the coaching brand associated with the service; its contact address is c/o Nuggt LLP at the address above. An AI conversation is not a live conversation with the coach behind coachparin and is not individually reviewed or approved by him.
2. Information we process
- Conversations: what you type or say, speech transcripts, the AI's replies, conversation titles, summaries, key takeaways and information needed to resume a session.
- Accounts and consent: your email address, authentication identifiers, account status and records of the terms version you accepted. Authentication services process the credentials you supply to them.
- Call and purchase records: call start/stop state, billable usage, minute balances, pack purchases, payment status, refund/dispute status and payment-provider references. Card details are entered on Stripe's hosted checkout; we do not receive or store your full card number or security code.
- Technical and support information: IP address and device/request information available to hosting and security services, authentication/session storage, operational errors and information you choose to send to support.
Please avoid trade secrets, privileged material, credentials, identification numbers and other people's personal information unless you are authorised to share it. Coaching conversations can contain sensitive information; share only what you are comfortable having processed for this service.
3. Voice, transcripts and automated processing
Your browser asks permission before using the microphone. Live microphone audio is sent to Deepgram for transcription. We do not store microphone recordings or generated speech audio in our application database or file storage. Text transcripts and AI replies are saved as described below. You can stop the microphone or end the call at any time.
Relevant conversation text, retained session context and retrieved coaching material are sent to Cerebras to generate answers and session summaries. Reply text is sent to Resemble AI to produce the synthetic voice. These are automated outputs and can be inaccurate.
When the speaking avatar is enabled, the AI's generated reply audio is sent to LiveAvatar by HeyGen to animate the avatar. LiveKit delivers the resulting avatar audio and video to your browser and processes connection/session metadata. Your microphone audio is not published to LiveAvatar or LiveKit by this application, and the application does not use your camera. The avatar is an AI presentation, not a live video connection to the coach.
When enabled, TypeSafe AI examines recent conversation text and relevant source excerpts to help decide whether a signup invitation is appropriate. This includes checking for useful context, a suitable moment, requests to stop or decline, and safety or sensitive situations. This is a limited product-flow decision, not a clinical assessment, employment decision or emotional profile used for advertising or personalised prices. You can decline the invitation.
4. Why we use information
We process information to provide the conversation and voice features, maintain accounts, save and resume sessions, calculate call time, take payments, handle refunds, record agreement to terms, respond to support requests, prevent abuse and meet legal obligations. We do not sell coaching conversations or use them for behavioural advertising. We do not use identifiable coaching conversations to train or fine-tune models.
Authorised personnel may access information when reasonably necessary for support, security, incident investigation, a complaint or a legal obligation. The service is not continuously monitored by a human, is not an emergency service and does not provide legally privileged communications. Provider retention and operational logs are separate from model training.
We measure unique identities that start calls, first verified accounts and active days within the last seven Singapore calendar dates. Article-level reports also measure completed coaching exchanges and return engagement. These reports use random account-linked identifiers, hashed identity and provider-session references, connection/activity dates and the article link that started a conversation. The private dashboard shows current account email addresses to authorised operators, but contains no coaching transcripts. Historical connection metadata can supplement retained application records without restoring deleted accounts or coaching content. Operator sign-in protection keeps only a purpose-bound IP hash for a short rate-limit window. We do not use a third-party advertising tracker or session replay for these measures.
5. Providers and international processing
Information is disclosed to the providers needed for the functions below. Provider infrastructure may process information outside Singapore, including in the United States. We do not promise that all processing stays in Singapore. We use access controls, encrypted transport and applicable contractual safeguards; no system is completely secure.
Cerebras - AI answers and summaries
Receives conversation/context text and reference excerpts. Its public privacy policy states that inference inputs and outputs are not retained, while service logs are kept as necessary; customer processing is also subject to the applicable service arrangements. Cerebras privacy policy.
Deepgram - live speech transcription
Receives microphone audio and returns transcripts. We set the per-request model-improvement opt-out. Deepgram states that opted-out audio and transcripts are retained only as needed to process the request; metadata and usage logs remain separate. Deepgram data handling.
Resemble AI - synthetic voice
Receives reply text and generates speech using the configured voice. Its policy states that synthesis text and generated audio are not retained after delivery. Voice-clone training recordings and models have a separate lifecycle; your microphone audio is not sent to Resemble to clone your voice. Resemble privacy policy.
LiveAvatar by HeyGen and LiveKit - speaking avatar
LiveAvatar receives generated reply audio to create the avatar's synchronized audiovisual stream. LiveKit transports that stream and processes connection and session metadata. Provider processing, operational logs and retention are governed by the applicable service arrangements; we do not promise a particular provider retention period. HeyGen privacy policy, LiveKit privacy policy and LiveKit data processing addendum.
TypeSafe AI - signup suitability
Receives the recent text/context used for the limited decision described above when this feature is enabled. Its policy states that input is not used to train or fine-tune models, and otherwise describes retention as reasonably necessary. TypeSafe privacy policy and data processing addendum.
Supabase - accounts and saved service data
Provides authentication and database storage for accounts, transcripts, summaries, consent, usage and purchase records. Retention follows our service lifecycle and deletion instructions, with provider backup and legal exceptions. Supabase privacy policy and customer data processing addendum.
Vercel - website and application hosting
Processes application requests and operational metadata for hosting, reliability and security. Logs and backups follow the applicable plan, configuration and retention requirements. Vercel privacy notice.
Stripe - payments
Processes checkout and payment information, fraud checks, receipts, refunds and disputes. Stripe may retain payment information after an account closes to meet legal, fraud-prevention, tax and accounting obligations. Stripe privacy policy.
6. How long information is kept
- Anonymous trial conversations: temporary records expire after 24 hours and are removed by routine cleanup, unless you save the conversation by creating an account before expiry.
- Product-growth counts: scheduled expiry of an anonymous trial detaches its account identifier and preserves content-free call and activity facts for historical reporting. These facts contain no email, IP address, transcript or article text. Some older pseudonymous connection records do not reveal whether their original identity expired or was explicitly deleted; we do not restore accounts or conversation content from those records. Explicit account or conversation deletion removes related analytics facts from the active database and records minimal hashed references to prevent their reimport.
- Saved account conversations: transcripts, replies and summaries remain until you delete the conversation or your account. Deleting a conversation removes its saved transcript and summary from the active application database.
- Account and service records: kept while needed to operate your account. When you delete your account, active conversation data and account links are removed; limited records may remain when necessary for security, dispute handling or legal obligations.
- Purchases and legal records: limited payment references, amounts, adjustments and records needed for legal obligations may be retained after account deletion. They do not need to include your coaching transcript. Stripe's own retention obligations also apply.
- Backups, logs and support: deletion from the active database is not immediate erasure from every backup or provider log. Restricted residual copies are retained only for the relevant backup cycle, security, support or legal need and are not used to resume deleted conversations.
7. Your choices and requests
You can stop a call, revoke microphone access in your browser, decline signup, delete saved conversations or request account deletion. Contact support@nuggt.io to request access, correction, deletion or withdrawal of consent, or to raise a privacy complaint. We may verify your identity before acting. We will explain any legal retention exception and respond in accordance with applicable law. Withdrawing information needed to operate a feature may mean that feature cannot continue.
The application uses browser session/authentication storage needed to keep you signed in and operate the service. You can clear it in browser settings, which may sign you out or remove access to an anonymous session. We do not use advertising tracking or session replay of your coaching conversation.
8. Age, changes and contact
The service is for adults aged 18 and over. If you believe a child has submitted personal information, contact us. We may update this notice as the service changes; the current version and effective date will appear here, and material changes will be communicated through reasonable means.
For privacy, support or legal notices: support@nuggt.io, Nuggt LLP, 39 Woodlands Close #08-61D, MEGA@WOODLANDS, Singapore 737856.