Privacy Policy

Spilki is operated by Limited Liability Company "AI" (LLC "AI"), registered in Ukraine (identification code 46045739), based in Kyiv, Ukraine ("Spilki", "we", "us").

For any privacy question or request, write to info@spilki.app .

Our two roles

• Controller. We decide why and how personal data is processed for our own customers (businesses and their team members who sign up for Spilki) and for visitors of spilki.app.
• Processor. When a business connects its Instagram, Telegram or website chat to Spilki, we process the conversations with its customers on behalf of that business. The business is the controller of its customers’ data and decides what the assistant does. We process this data only to provide the service, as set out in section 11 of our Public Offer .

If you wrote to a business that uses Spilki and have a question about your data, please contact that business first. We will help it handle your request.

Spilki customers and their team members

• Account data: name, email address, profile photo and sign-in details that you provide when you register or that come from the sign-in method you choose (for example, a social login).
• Organisation data: business name, settings, team members and their roles, connected channels and integrations.
• Content you add to set up the assistant: instructions, knowledge-base documents and files, website pages you ask us to import, product catalogues, prices, services and working hours.
• Billing data: plan, subscription status, payment history and invoices. Card payments are made on the payment page of monobank (JSC "Universal Bank"); we never receive or store your full card number.
• Support requests: messages you send us and our replies.

Customers of businesses that use Spilki

When you write to a business through a channel connected to Spilki, we process on behalf of that business:

• Messages, comments and attachments you send (text, photos, files, voice messages), and the replies of the assistant and the business team.
• Profile data the channel provides: your name, username, account identifier and profile photo on Instagram or Telegram.
• Details you share in the conversation or that the business adds: phone number, email address, delivery address, orders, bookings and notes.
• Website chat: the visitor identifier and the conversation history. If the website passes the details of a signed-in visitor to the chat, also their name, email address and phone number.

Visitors of spilki.app

• Google Analytics 4 data, only if you accept analytics cookies. See our Cookie Policy .

Technical data

• IP address, browser and device type, pages and features used, timestamps and error reports. Server logs and monitoring tools collect them to keep the service working and secure.

Data that comes from services a business connects (channels, booking systems, stores, Google services and others) is described in section 10 .

• To provide the service: receive and send messages in connected channels, generate the assistant’s replies, create orders and bookings, notify the business team and show conversation history in the app.
• To run your account: registration, sign-in, team management, payments, invoices and support.
• To keep Spilki secure and reliable: prevent abuse and fraud, monitor performance, investigate errors and incidents.
• To improve Spilki: understand which features are used, relying on aggregated or de-identified data where possible.
• To communicate with you: service and billing notices, and product news that you can unsubscribe from at any time.
• To comply with the law: accounting, tax and responses to lawful requests from authorities.

We do not sell personal data and do not use it for targeted advertising.

The Spilki assistant runs on large language models from OpenAI. To answer a message, we send OpenAI, through its API, only what that answer needs: the recent conversation, relevant parts of the business’s knowledge base and settings and, for bookings, the free time slots we calculated. We also use OpenAI to transcribe voice messages, understand images and build the search index of the knowledge base.

• OpenAI acts as our processor. Under its API terms, data sent through the API is not used to train OpenAI’s models. OpenAI may keep it for a limited time (up to 30 days) to detect abuse and then deletes it.
• We do not use your content or your customers’ conversations to train our own or third-party AI models.
• AI replies can contain mistakes. The business sees every conversation in the app, can correct the assistant and can take over a conversation at any time.
• The assistant does not make decisions that have legal or similarly significant effects on people.

We process personal data under the Law of Ukraine "On Personal Data Protection" and, where it applies, the EU General Data Protection Regulation (GDPR), on these grounds:

• Contract: to provide Spilki under our Public Offer and to act on your requests before you sign up.
• Legitimate interests: to protect the service, prevent abuse, improve Spilki and keep business customers informed about the service. We weigh these interests against your rights.
• Consent: for analytics cookies on spilki.app, for connecting optional integrations such as Google, and for marketing emails where the law requires it. You can withdraw consent at any time.
• Legal obligation: for accounting, tax and responses to lawful requests.

For conversations we process on behalf of a business, the business is responsible for having a legal basis to process its customers’ data.

We share personal data only with service providers that help us run Spilki (processors), under contracts that require them to protect it and use it only on our instructions:

• Amazon Web Services (AWS), EU (Frankfurt) region: hosting, database, file storage, email and content delivery.
• OpenAI (USA): AI models for replies, voice transcription, image understanding and knowledge-base search.
• Clerk (USA): sign-in and account management.
• Novu (USA): in-app and email notifications to the business team.
• Sentry: error reports from the Spilki app (personal data and on-screen text are masked).
• Grafana Cloud (United Kingdom): technical monitoring of the service.
• Firecrawl (USA): importing website pages that a business adds to its knowledge base.
• monobank (JSC "Universal Bank", Ukraine): card payments for subscriptions.
• Google (Google Analytics 4): analytics on spilki.app, only with your consent.

We also exchange data with services that a business chooses to connect, so that those connections work: Meta (Instagram), Telegram, Google Calendar and Google Drive, and business integrations such as Beauty.Pro, Altegio, Choice, WooCommerce and Prom.ua. These services process data under their own privacy policies.

We may disclose data when the law or a lawful request of a public authority requires it; to protect the rights and safety of Spilki, our customers or others; or as part of a merger, acquisition or sale of assets, in which case this policy continues to apply to your data.

Our main infrastructure is in the European Union (AWS, Frankfurt). Some of the providers listed above process data in the USA or the United Kingdom. When data leaves Ukraine or the European Economic Area, we rely on appropriate safeguards, such as the European Commission’s Standard Contractual Clauses included in our providers’ data processing terms, or on an adequacy decision.

• Account and organisation data: while the account is active. When an organisation is deleted, we immediately disconnect its channels and integrations, revoke and delete its integration access tokens and delete its knowledge-base files. The remaining organisation and conversation records are taken out of use at once and permanently deleted 30 days after the organisation is deleted. Aggregated statistics that contain no personal data (for example, the number of conversations per day or AI usage and costs) may be kept for reporting and accounting.
• Conversations: while the organisation is active, or until the business deletes them.
• Website chat: visitor data is anonymised after 365 days without activity. The conversation history in the visitor’s browser stays in local storage until the visitor clears it.
• Instagram and Telegram profile photos are cached for up to 90 days.
• Billing and accounting records: for the period required by Ukrainian tax and accounting law.
• Server logs: up to 30 days. Database backups: up to 7 days.
• Website analytics: 14 months (see the Cookie Policy ).
• Support correspondence: as long as needed to resolve the request and for our legitimate record-keeping.

• All connections to Spilki are encrypted with TLS.
• Databases and file storage are encrypted at rest; access tokens for connected channels and integrations are additionally encrypted at application level.
• Each organisation’s data is isolated at database level (row-level security).
• Only staff who need it have access to production systems, and secrets are kept in a dedicated secrets store.
• Card data is handled by monobank and never reaches our servers.

No system is perfectly secure. If a personal data breach affects you, we will notify you and the competent authorities as the law requires.

A business can connect third-party services to Spilki: messaging channels, booking and reservation systems, online stores and catalogues, knowledge-base sources, Google services and other services such as weather forecasts. The business itself turns each connection on in the Spilki app and can turn it off at any time. Without such a connection we receive no data from that service.

How we handle integration data

• Access. We connect through the service’s official mechanism: a sign-in where you grant permission (OAuth), or an API key, token or other connection details that the business enters itself. We ask only for the permissions, and read only the data, that the feature the business enabled cannot work without.
• Purpose. Integration data is used only to make that feature work for the business that connected it: so the assistant can receive and send messages, see free times and create bookings, answer questions about products, prices and stock, or rely on the knowledge base.
• Protection. Tokens, API keys and other integration credentials are encrypted at application level. Each organisation’s data is isolated at database level (row-level security) and hosted on Amazon Web Services in the EU (Frankfurt).
• Sharing. We do not sell integration data and do not use it for advertising, including retargeting, personalised or interest-based ads. What the assistant produces (for example, a booking or an order with the customer’s name and phone number) is sent to the connected service of the same business. Otherwise this data goes only to processors the feature cannot work without (section 6), including OpenAI, which receives only what the feature needs: content to build the knowledge-base search index and, for each reply, only what that reply needs (see section 4 ), or to authorities when the law requires it.
• AI. We do not use integration data, or data derived from it, to train our own or third-party AI models.
• Retention and deletion. We keep only what the feature needs: connection settings, synced catalogues or menus, imported documents. When a business removes an integration or channel in Spilki, we stop calling the service and delete the stored credentials of that connection. When an organisation is deleted, the periods in section 8 apply. For a Google account used by several features, the credentials are deleted when the last of them is disconnected (see Google services ).

Messaging channels

Instagram (via Meta), Telegram, the website chat and Prom.ua chats. We receive the messages, comments, attachments and profile data the channel provides, and send replies on behalf of the business (see section 2 ). If the business adds its Telegram bot to a group, we store that group’s messages so the assistant can follow the conversation. If the business asks, we also import its Instagram posts into the knowledge base.

Booking and reservation systems

Altegio, Beauty.Pro and Choice (menus and table reservations). We read services, staff, menus, free times and tables, and create bookings or reservations (and, where the system supports it, cancel them) with the name and contact details the customer gave in the conversation.

Online stores and catalogues

WooCommerce and Prom.ua. We read products, prices, stock, delivery and payment options, and the orders and buyer details needed to find an order, so the assistant can answer questions about products and order status. We do not change products or create orders in these systems.

Payments

monobank is used only for paying the Spilki subscription (see sections 2 and 6). Card data is handled by monobank and never reaches us.

Knowledge-base sources

Files the business uploads; website pages it asks us to import (through Firecrawl, refreshed automatically at most once a day); and Google Docs, Google Sheets and Google Drive files by link. We read Google documents through a Spilki service account with read-only access, and only those the business has made available by link or shared with that account. A business’s own Google account is not needed for link imports. A business can also pick files from Google Drive with its connected Google account; we then read only those files, with the access described under Google services , and keep them up to date.

Other services

OpenWeather: the assistant can look up the weather for a given city. The business connects its own API key, and we send the service only the place the forecast is needed for.

Google services

A business can connect its Google account to Spilki so the assistant can answer customers from Google Drive files picked for the knowledge base, book appointments in Google Calendar and send files from Google Drive to the business’s customers. One connected account can serve several of these features; for each one we ask only for the permissions it still lacks. We access Google user data only after the account owner grants permission on Google’s consent screen, and only for these purposes:

• Basic profile (openid, email, profile): to identify the connected Google account and show in Spilki which account is connected.
• Google Calendar list (calendar.calendarlist.readonly): to let you choose which of your calendars receives bookings. We store only the selected calendar’s ID and name.
• Calendar availability (calendar.events.freebusy): to read busy time ranges (start and end only, not event details) in your booking calendar, so the assistant offers customers only free times within your business hours. Availability is used at the moment of the request and is not stored.
• Events on calendars you own (calendar.events.owned): to create an event for each booking, to move or delete that event when the customer reschedules or cancels, and to check whether you moved or deleted a Spilki-created event in Google Calendar, so the booking stays in sync. Spilki reads and changes only events it created.
• Google Drive files you pick for the knowledge base (drive.file): to read the content of the files you explicitly select in the Google file picker (for example, documents, spreadsheets, presentations and PDFs), store it in the business’s knowledge base and build its search index, so the assistant can answer your customers from it. Our AI provider processes this content to build the search index and, for each reply, receives the relevant parts of it, as described in section 4 ; it does not use them to train its models. We re-check picked files about every 15 minutes, and when you ask, so the knowledge base reflects the changes you make in Google Drive.
• Google Drive files you pick for sending (drive.file): to access only the photos and documents you explicitly select in the Google file picker, so the assistant can send them to your customers on request. Selected files are cached in our storage so they can be sent as chat attachments. Photos are also described by our AI provider (see section 4 ), and we store these descriptions and their search index so the assistant can find the right file for a customer’s request.

The drive.file permission gives us no access to any other files in your Google Drive, including the other files of a folder or shared drive you browse in the picker.

Everything above in this section applies to Google user data: OAuth tokens are encrypted (AES-256), and we do not sell this data, do not use it for advertising, and do not use it, or data derived from it, to develop, improve or train generalised or non-personalised AI or machine-learning models.

Spilki staff do not read Google user data unless you give us explicit permission for a specific case (for example, a support request), it is necessary for security purposes such as investigating abuse, or it is required by law.

You can disconnect Google at any time in the Spilki app, for one feature or for the whole account. When you disconnect the knowledge base, we stop reading the picked files; the content already imported stays in the knowledge base until you delete it there, and is deleted with the organisation within the periods in section 8 . When you disconnect the last feature on an account, we revoke our access at Google and delete the stored tokens immediately. You can also remove Spilki’s access at myaccount.google.com/connections .

Spilki's use and transfer to any other app of information received from Google APIs will adhere to Google API Services User Data Policy , including the Limited Use requirements.

Questions about integration data, including Google user data: info@spilki.app .

Depending on where you live, you have the right to:

• know what personal data we hold about you, where it comes from, why and where it is processed, and get a copy of it;
• have inaccurate data corrected;
• have your data deleted;
• restrict or object to processing, including objecting to direct marketing at any time;
• receive your data in a structured, machine-readable format and have it transferred to another service;
• withdraw consent at any time, without affecting processing that took place before;
• lodge a complaint with a data protection authority.

These rights are provided by Article 8 of the Law of Ukraine "On Personal Data Protection" and, for people in the European Economic Area, also by the GDPR. In Ukraine, complaints are handled by the Ukrainian Parliament Commissioner for Human Rights; in the EEA, by your local data protection authority. We would appreciate the chance to resolve your concern first.

To exercise your rights, write to info@spilki.app . We respond free of charge within 30 days and may ask you to confirm your identity. If your data is processed on behalf of a business that uses Spilki, we will forward your request to that business and help it respond.

• spilki.app uses Google Analytics cookies only if you accept them in the cookie banner. Details are in the Cookie Policy .
• The Spilki app (home.spilki.app) uses cookies and local storage that are strictly necessary for sign-in and for keeping your session secure.
• The website chat does not set cookies. It keeps the visitor’s session identifier and conversation history in the browser’s local storage, so the conversation continues on the next visit.

Spilki is a business service and is not intended for people under 16. We do not knowingly collect children’s personal data for our own purposes. If you believe a child has given us personal data, contact us and we will delete it. Businesses that use Spilki are responsible for their communication with customers of any age.

We may update this policy when Spilki changes or the law requires it. We publish each new version on this page with a new date. We will notify customers of significant changes by email or in the app before they take effect.

This policy is published in Ukrainian and English. The Ukrainian version is the original; if the versions differ, the Ukrainian version prevails.

• Limited Liability Company "AI" (Spilki)
• Identification code (USREOU): 46045739
• Location: Kyiv, Ukraine
• Email for privacy requests: info@spilki.app
• Support: support@spilki.app