Security & Trust
The claims on our homepage have a page behind them. Here is how Avrosh keeps data safe: strict per-business isolation in the database, a customer surface that collects nothing personal, an AI that cannot make things up, three data regions you can be assigned to, a sub-processor list generated from the running configuration rather than typed out, and no advertising trackers anywhere.
Isolated per business
Avrosh is multi-tenant on PostgreSQL with row-level security enforced at the database layer. Every query runs under the tenant that issued it, so one business can never read another business's rows. Isolation is a property of the database, not just the application.
A zero-PII customer surface
The QR chat carries no account, no phone number and no login. A customer starts a conversation with nothing personal collected, so there is little to leak and nothing to consent-manage on that surface.
Data residency, in three regions
Avrosh runs in three regions. Your business is assigned to one of them, and your application servers and database live there.
- United States, Northern Virginia. For customers in the Americas.
- Singapore, Singapore. For customers across Asia Pacific.
- European Union, Frankfurt. For customers in Europe.
A region decides where your servers and your database live. It does not yet decide where the AI runs, and we would rather you heard that here.
Today every model certified to serve an account is hosted outside the European Union: the language models and the knowledge base search are served from Singapore, speech to text from the United States, and our speech vendor does not disclose an origin. EU-served models are in the catalogue but are not yet certified, so they cannot be assigned to a business. If in-region AI processing is a requirement for you, tell us before you sign rather than after, and we will tell you where certification stands.
Encryption
Traffic is encrypted in transit with TLS. Data at rest is protected by the hosting platform.
The truth firewall, and where it is thin
Avrosh answers only from each business's own data, and it escalates to a human when it is unsure. Treating a wrong answer as worse than no answer is a security property: it protects customers and operators from misrepresentation.
The mechanical part is narrower than the slogan, so here is the real scope. Prices, clock times, durations, arrival estimates and calendar dates in a reply have to appear in what a tool actually returned on that turn, or the reply does not go out. That check is real and it runs on every answer.
Names are checked too, but only on the accommodation path, where a room type invented out of nothing is caught. The equivalent check for a salon service, a dental procedure or a menu item is not built yet, and neither is name checking outside English. We would rather you knew which half is enforced by code than take the sentence on faith.
No certifications, and no pretending
We hold no SOC 2 report, no ISO 27001 certificate, no HIPAA status and no PCI attestation. Competitors print those logos; we do not have them, and a procurement team finds that out in the first call.
What we can give you instead is a written answer to your security questionnaire, the per-model provenance record above, and the reasoning behind every claim on this page. Card data never reaches us at all: subscriptions run through a payments provider acting as merchant of record, so we hold a subscription status rather than a card number.
No advertising trackers
No Meta pixel, no ad networks and no third-party ad trackers, on this site or in any customer chat.
Data subject rights, built in
For a customer who holds an account with a business, export and permanent erasure are built into the product and are a click rather than a support ticket. An export returns their contact record, bookings, waitlist entries and consent history. An erasure strips every identifier in place while leaving the booking rows the business needs for its own records, and where branches are grouped it leaves no identifying bytes at any of them.
An operator-side export of your own business content is not self-serve yet. Ask us and we provide a copy. This page said otherwise until we checked, which is the kind of claim a buyer tests.
AI sub-processors
These are the models that can actually be assigned to a business today. The table is generated from the same configuration the running system picks from, so it changes when the product changes rather than when someone remembers to edit this page.
Two facts are tracked separately on purpose. Served from is where the request is actually answered. Vendor jurisdiction is where that company is incorporated. They are often not the same place, and a page that prints only one of them is not answering the question your counsel is asking. Where we have not verified an origin ourselves, the last column says so instead of leaving a blank.
| Role | Provider | Served from | Vendor jurisdiction |
|---|---|---|---|
| Embeddings for knowledge base search | Alibaba Cloud (DashScope) | SingaporeStated by the vendor | Alibaba Cloud (jurisdiction unverified) |
| Language model, chat | Alibaba Cloud (DashScope) | SingaporeNot disclosed | Not recorded |
| Language model, classification | Alibaba Cloud (DashScope) | SingaporeNot disclosed | Not recorded |
| Language model, voice | Alibaba Cloud (DashScope) | SingaporeNot disclosed | Not recorded |
| Speech to text | Soniox | United StatesMeasured by us | United States |
| Text to speech | Fish Audio | Multiple regionsNot disclosed | unverified |
In the catalogue, not yet certified
Avrosh is built to be multi-provider, and these are the vendors we hold relationships with but have not certified. A model is only offered once we have measured it and built an adapter for it, so nothing here touches anyone's data. It is published so that a question like "can you serve my language model from the European Union" gets an honest answer with a reason rather than a flat no.
| Role | Provider | Would serve from | Vendor jurisdiction |
|---|---|---|---|
| Language model, chat | DeepSeek | Multiple regions | China |
| Language model, chat | Mistral | European Union | France |
| Language model, chat | OpenAI | United States | United States |
| Language model, classification | OpenAI | United States | United States |
| Language model, voice | Mistral | European Union | France |
| Language model, voice | OpenAI | United States | United States |
| Speech to text | Alibaba Cloud (DashScope) | Singapore | Alibaba Cloud (jurisdiction unverified) |
| Speech to text | Deepgram | United States | United States |
| Speech to text | Speechmatics | European Union | United Kingdom |
| Text to speech | bland | United States | United States |
| Text to speech | Cartesia | United States | United States |
| Text to speech | ElevenLabs | United States | United States |
| Text to speech | MiniMax | Multiple regions | unverified |
| Text to speech | OpenAI | United States | United States |
Platform sub-processors
Beyond the AI layer, a small set of vendors runs the service. These are listed by role; the named vendor for each is confirmed on request at hello@avrosh.com, because publishing a guess would be worse than publishing nothing.
| Role | Purpose | Region |
|---|---|---|
| Hosting | Application servers and the PostgreSQL database | The region you are assigned: Virginia, Singapore or Frankfurt |
| Payments | Subscription billing, as merchant of record | Global payment processor |
| Transactional email such as receipts and notifications | Email delivery provider | |
| Telephony carrier | Connecting phone calls to and from the AI | Carrier network, by number country |
Reporting a vulnerability
Found a security issue? Write to hello@avrosh.com with the details. We investigate every report and will acknowledge you.
This page reflects how Avrosh operates today and is reviewed regularly. For a security questionnaire or the named sub-processor list, write to hello@avrosh.com.