HAMI Voice is operated by NIRST LTD, a company registered in Scotland under company number SC710095, with registered office at 4 Rubislaw Terrace, Aberdeen, AB10 1XE, United Kingdom. In this Privacy Policy "we", "us", and "our" refer to NIRST LTD. We are the data controllers responsible for personal data collected and processed in relation to the HAMI Voice platform and services. We are registered with the UK Information Commissioner's Office (ICO) under registration number ZB728752.
The privacy and security of your data is important to us. Please review this Privacy Policy to understand how we handle your information. By using or accessing the HAMI Voice platform you acknowledge and accept the practices described below.
Last updated: 2026-10-02
We collect and process personal data in compliance with applicable data protection laws, including the UK General Data Protection Regulation (UK GDPR) and the EU GDPR.
When you sign up for HAMI Voice we collect:
Your account is associated with a team. Your team owner may also see your name and email address as part of team management.
Where you create or administer a team you may also provide:
When you create or run surveys on HAMI Voice you provide content, including:
You retain ownership of this content. We process it solely to provide the service to you.
When you choose to use the voice-driven survey mode, your speech is recorded in real-time and converted to text by a third-party speech-to-text service (Soniox). Real-time audio is streamed directly to the speech-to-text provider and is not stored by default. Soniox does not retain audio after processing and does not use your data to train its models.
While voice mode is active, recording continues when the HAMI Voice app is in the background — for example when your screen is locked, when you switch to another app, or when your phone is in a pocket or pouch — so you can complete a survey hands-free. Your device's operating system displays a microphone indicator (a small dot in the iOS status bar, a notification icon on Android) whenever the microphone is active. Recording stops when you exit voice mode, end the survey session, or close the app.
If your team has enabled the optional voice audio log, the platform additionally stores per-utterance audio clips and transcripts for up to 14 days. The audio log is disabled by default and only available where a team administrator opts in. When enabled, all users in the team are shown a consent disclosure before voice mode begins ("This inspection will be recorded for quality and debugging purposes. Recordings are automatically deleted after 14 days."). You can withdraw consent at any time by declining the consent prompt or by switching to manual text entry.
Audio log recordings are stored on AWS S3 in the eu-central-1 region (Frankfurt, Germany) and are automatically deleted by S3 lifecycle rule after 14 days. Transcripts and AI extraction metadata associated with utterances are kept with the survey session. They are removed when the session is deleted (see "Deleting a survey session"), when your account is deleted, or when you ask us.
Voice features are entirely optional. You may always use manual text entry instead. Your voice data is not used to identify who you are, and never to log you in.
When the start-phrase feature is enabled, HAMI Voice distinguishes your voice from other voices within earshot, so that speech from bystanders is not written into your survey as if you had said it. You say a short start phrase, and the speech-to-text service groups your later speech with it under a temporary label that is discarded when the connection closes.
Where replay is also enabled, the web app keeps about 1.5 seconds of you saying the start phrase, so you are not asked for it again each time the survey reconnects. That recording stays in your browser, on your device, and is never stored by us or by any third party. Each time the survey reconnects, it is played to the speech-to-text service together with your live speech, and the service does not retain it. It is deleted when you complete the survey in that browser. If you leave the survey unfinished, it stops being used after 7 days and is deleted the next time you open a voice survey in that browser. Clearing your browser's site data deletes it at any time. The mobile app keeps no such recording.
The recording and the temporary labels are used only to match your voice within one survey, and never to identify you as a person or to authenticate you. No voiceprint is created: the recording is ordinary audio.
Where you capture photographs as part of a survey session, the image files are uploaded to AWS S3 in the eu-central-1 region (Frankfurt, Germany). Images are linked to the survey session, group instance, or field that they were captured against. Deleting a photo in the app deletes its stored file too. Otherwise the file is removed when the session or your account is deleted, or on request.
Where a survey field requests a signature, the signature is captured on-screen and stored as a small image embedded directly in the survey session record. Signatures are used to evidence acceptance or sign-off of a completed report and are not used for biometric identification or authentication. Signatures are removed when the session or your account is deleted, or on request.
You can delete a survey session from its page in the web app. Your team's owner can delete any session. A team member can delete an unfinished session assigned to them. Completed and archived sessions can be deleted only by the team owner. Every delete asks you to type the survey's title to confirm. A deleted session disappears from HAMI Voice at once. A phone that already holds a copy can no longer save changes to it or send it back. We keep it for 30 days in case it was deleted by mistake; within that time you can ask us to restore it. After 30 days it's removed for good, with its photos, signatures, voice recordings, transcripts and stored reports. It leaves our database backups up to 30 days after that.
Survey reports are generated from your data when you request them. The generated report and its preview are stored on AWS S3 in the eu-central-1 region (Frankfurt, Germany) so that they can be downloaded again and shared without regenerating them. Stored reports are kept with their survey session and removed when it is deleted (see "Deleting a survey session"), or sooner if you ask us.
We collect operational telemetry that helps us run and improve the service:
We use both session cookies (which expire when you close your browser) and persistent cookies (which remain on your device until they expire or you delete them) to authenticate your session, remember your preferences, and analyse usage. You can control cookies through your browser settings; disabling cookies may limit your ability to use the platform.
The information collected is used for the purpose of providing and improving the platform. We may use your personal data to:
| Processing activity | Lawful basis (UK GDPR Art. 6) |
|---|---|
| Account creation and authentication | (b) Performance of a contract |
| Storing and processing survey content you create | (b) Performance of a contract |
| Voice recording and transcription (real-time) | (a) Consent — voice mode is optional |
| Voice audio log (when team has opted in) | (a) Consent — admin-enabled, per-user disclosure |
| Telling your voice apart from others during a survey (start phrase) | (a) Consent — the feature is optional |
| Separating bystander speech so it is not recorded as yours | (f) Legitimate interest — to keep other people's words out of your survey |
| AI processing of survey input | (b) Performance of a contract |
| AI Template Builder processing of uploaded documents | (b) Performance of a contract |
| Service emails (security, account, billing) | (b) Performance of a contract |
| Marketing communications | (a) Consent (opt-in, opt-out at any time) |
| Aggregated analytics and service improvement | (f) Legitimate interest |
HAMI Voice includes AI-powered features that process your input using third-party AI services:
We use AI-assisted features on the basis of contract performance (delivering the service you have signed up for) and, for voice, consent. You are informed when interacting with an AI-powered feature.
We do not use customer data to train any AI models, and our AI sub-processors have committed not to use your data for their own model training.
AI transparency (EU AI Act). Our AI features are designed to meet the transparency requirements of Article 50 of the EU AI Act: you are always told when you are interacting with an AI feature, and AI-generated content in exported reports (the executive summary) is visibly marked as generated with AI assistance. HAMI Voice contains no prohibited AI practices — no emotion recognition, no biometric identification or categorisation, and no social scoring — and no high-risk AI systems as defined by the EU AI Act. AI outputs are assistive drafts: a human always reviews and confirms them, and no AI feature makes automated decisions that produce legal or similarly significant effects about you.
We use the following third-party services to deliver the platform. Each acts as a data processor on our behalf under appropriate data processing agreements:
| Provider | Purpose | Data processed | Data location |
|---|---|---|---|
| Vercel | Web application hosting and page-view analytics (Vercel Web Analytics, which sets no cookies and stores no personal identifiers) | Application data, request logs, anonymised page-view counts | EU region; Vercel Inc. is a US company (see "International Transfers") |
| Neon | PostgreSQL database services | All stored personal data | EU region; Neon is part of Databricks, Inc., a US company (see "International Transfers") |
| AWS (S3) | File storage (photos, signatures, audio log, template source documents) | Image files, audio files, document uploads | Frankfurt, Germany (eu-central-1) |
| AWS (Bedrock) | AI inference (PDF document analysis) | Uploaded PDF source documents | Frankfurt, Germany (eu-central-1) |
| xAI | AI inference (survey assistant, template builder default) | Text prompts, transcribed speech, image content of uploaded forms | EU endpoint (eu-west-1); xAI is a US company (see "International Transfers") |
| Soniox | Real-time speech-to-text transcription | Audio streams (real-time, zero retention) | EU processing endpoint; Soniox is a US company (see "International Transfers") |
| AWS (SES) | Transactional email delivery (account verification, password reset, account-deletion notices, trial-ending reminders) | Email address, recipient name, message content | Frankfurt, Germany (eu-central-1) |
| Stripe | Subscription billing and payment processing. Card details are entered on Stripe's own pages and never reach our servers. | Name, email address, billing address, payment method, invoices | Stripe Payments Europe Ltd (Ireland); Stripe may process in the United States (see "International Transfers") |
Optional providers (active only where configured by us as alternates to the defaults above):
| Provider | Purpose | Data location |
|---|---|---|
| Google Cloud (Vertex AI, Gemini) | AI inference | Belgium (EU) |
| OpenAI | AI inference | OpenAI OpCo, LLC (US) for UK customers; OpenAI Ireland Ltd for EEA customers |
| Mistral AI | AI inference, OCR | EU (Paris, France) |
These providers process your data solely for the purpose of delivering the relevant feature and do not use your data for their own purposes, including model training.
Several of the sub-processors listed above are US companies, even where they process your data on servers in the EU: Vercel, Neon (part of Databricks), Soniox, xAI and, where configured, OpenAI. Your stored data, audio and transcripts are processed in EU regions wherever the provider offers one. Stripe may process payment data in the United States. These transfers take place under the EU-U.S. Data Privacy Framework and its UK Extension (Vercel, Neon), or under the EU Standard Contractual Clauses with the UK Addendum or UK International Data Transfer Agreement (Soniox, xAI, OpenAI), or an equivalent transfer mechanism (Stripe). AWS, Google Cloud and Mistral AI contract with us through European companies. We do not transfer your personal data outside the UK or EEA without such a mechanism in place.
We retain different types of data for different periods:
| Data type | Retention period |
|---|---|
| Account and profile data | Duration of your account, plus any legal retention period |
| Team and brand data | Duration of your team's account; a replaced or removed logo file is deleted at once |
| Survey templates | Until deleted by your team, with their reference images |
| Survey sessions and responses | Until you delete the session or your account, or on request. Archiving a session hides it but keeps its data |
| Deleted survey sessions | Removed for good 30 days after the delete, with all their files; out of database backups up to 30 days after that |
| Survey photos | Kept with the session; removed when you delete the photo, the session or your account, or on request |
| Signatures | Kept with the session; removed when the session or account is deleted, or on request |
| Voice audio log recordings (when team has opted in) | 14 days (auto-deleted by S3 lifecycle rule) |
| Voice audio log transcripts and AI extraction metadata | Kept with the session; removed when the session or account is deleted, or on request |
| Generated reports | Kept with the session; removed when the session or your account is deleted, or on request |
| Report previews | 7 days (auto-deleted by S3 lifecycle rule) |
| AI Template Builder source documents | 7 days after the template is saved or abandoned; 30 days after you last worked on an unfinished one |
| Expired sign-in sessions and one-time codes | Deleted within a day of expiring |
| Application logs (telemetry) | Up to 30 days (mobile app logs: 7–14 days) |
| Database backups | Up to 30 days |
We implement comprehensive security measures to protect your personal data:
Under data protection law you have the following rights in respect of your personal data:
To request deletion of specific data without closing your account, email us at support@hami-voice.com describing the data you would like removed (for example, a particular survey session, voice audio log recording, stored report, or uploaded template source document). We will action your request within 30 days, except where we have an overriding legal obligation to retain the data (for example, records required for tax, audit, or regulatory purposes).
To delete your account and all associated personal data, follow the steps on our Account and Data Deletion page. That page also describes how to request deletion if you can no longer sign in.
In order to process any of these requests we may need to verify your identity for your security. To exercise any of these rights, contact us at support@hami-voice.com.
We will only send you marketing communications about HAMI Voice with your consent. You can withdraw consent at any time by clicking the unsubscribe link in any marketing email or by contacting us at support@hami-voice.com. Withdrawing consent for marketing does not affect service emails (account notices, security alerts, billing if applicable).
HAMI Voice is intended for use by professionals in a business context. The platform is not directed at children, and we do not knowingly collect personal data from anyone under the age of 16. If you believe we have inadvertently collected such data, please contact us so we can delete it.
We reserve the right to update this Privacy Policy. We will post any changes on this page with a revised "Last updated" date. Material changes will be communicated to existing users by email or in-app notice.
If you have any questions or concerns about this Privacy Policy or our data practices, please contact us at support@hami-voice.com.
For postal correspondence:
NIRST LTD 4 Rubislaw Terrace Aberdeen, AB10 1XE United Kingdom