Privacy Policy (last updated 11/18/2024)
At Thunk.AI, we care about and respect the privacy of your data! This privacy policy explains what data is recorded in each category and how this data is used.
Policy Updates: The Thunk.AI service is evolving rapidly. As the Thunk.AI service/product evolves, we may make changes to our privacy policy and reserve the right to do so. If we change this privacy policy, we will make a note of it on our site. Your continued use of the Thunk.AI service indicates that you agree to our changes. You can see the most recent version of the policy by looking at the date in the heading of the policy.
Personal Information: You provide personally identifiable information when you sign up for the Thunk.AI service. When you sign up via a third-party authentication provider (like Google), the Thunk.AI service retrieves and saves some information about you from the provider (the specific information is controlled by the authentication provider and by the permissions you approve when you sign up). In particular, the Thunk.AI service receives your email address. In order to interact with your data in your linked account (such as Google Drive), our backend service also maintains authentication (OAuth) tokens that allow Thunk.AI to act "on your behalf" to read and write data. These tokens are securely stored along with your personal information. Your personally identifiable information will not be sold or rented to third parties in a way that identifies you as an individual. As we develop our business, we may buy or sell assets, and, depending upon the transaction, your personally identifiable information may be one of the transferred assets. In the event that we are acquired by another company, your personally identifiable information may be part of the assets transferred to the acquiring party. We may share anonymized aggregated order information with third parties as a way to understand, explain and analyze our business (eg: X% of our users utilized feature Y).
Application Data: The Thunk.AI service stores and maintains: (a) project-related data including relevant files and email data, (b) messages between users and their Thunk AI agents, and (c) messages between the Thunk AI agents and AI models (like OpenAI’s GPT). Some data used by your Thunk projects may originally reside in a third-party storage provider (eg: Google Drive). For effective processing, a copy or extract or index of the data may be maintained in the Thunk.AI service or its subprocessors.
Usage Data: the Thunk.AI service uses browser session cookies and browser local storage to maintain user session state. This helps to speed up user login. You can set your browser to disable cookies and local storage, but this will significantly compromise or disable product usability. As you use the Thunk.AI service, it logs usage information in backend log files and logging and debugging services, including the features you accessed on our site, and standard browser-provided information. We utilize this information not only to debug issues and problems, but also to learn aggregated patterns across all our users and thereby improve the Thunk.AI service. The Thunk.AI service may send notifications via email and other notification mechanisms. We may employ clear gifs to track the effectiveness of our emails. We may share anonymized non-personally identifiable usage data with third-party partners in order to understand, analyze, or explain aggregate customer usage of the Thunk.AI service. We may use multiple third-party products and services like Mixpanel to record and analyze usage data on our site. This tracking of usage patterns is essential for the effective operation and improvement of our platform. Some browsers have incorporated "Do Not Track" (DNT) features that can send a signal to the websites you visit indicating you do not wish to be “tracked”. Because there is not yet a common understanding of how to interpret the DNT signal, Thunk.AI services do not currently respond to browser DNT signals. This is similar to the behavior of services from other large and established software providers, while the online industry works towards standardizing the meaning of this signal.
Data Security and Durability: We store user personal information and access tokens in databases hosted by cloud-computing hosting providers. We take reasonable technical steps to protect your information including the use of passwords, OAuth, secure HTTP, etc. Data is encrypted at rest and encrypted in transit. However, we cannot guarantee that your information is immune to a breach in the security and privacy mechanisms implemented by us and the hosting providers.
Sub-Processors: The Thunk.AI service utilizes the following sub-processors in order to provide users with an effective service. Each of these services is governed by its own terms of service and its contractual obligations to Thunk.AI.
AI services
Open AI
Cloud hosting services
Google Cloud
Cloudflare
Logging and Monitoring services
Helicone.ai
Sentry
Mixpanel
Jsreport
Serper
Operations
DynamicPDF
We may update this list from time to time as our service evolves.
Data Access and Data Retention:
Your Files and Emails: Your AI uses (reads and indexes) the files and emails that it has access to in order to find and synthesize relevant information when needed by your projects. Your AI only accesses files for the purpose of or during the execution of project work. Your AI also generates new files or images when needed by your projects, writing them to the Thunk.AI folder in your cloud file system or in other locations as appropriate.
Storage on Google Cloud: The Thunk.AI service is hosted on Google Cloud. All data associated with your thunks in the Thunk.AI service is stored on Thunk.AI’s tenant in Google Cloud. This includes: (a) project-related data including relevant files and email data, (b) messages between users and their AI agents, (c) messages between the AI agents and AI models (like OpenAI’s GPT), (d) execution and monitoring logs, (e) account-related data. All your data is encrypted-at-rest and encrypted-in-transit.
If you delete your Thunk.AI account, the Thunk.AI service will delete all data in categories (a), (b), (c) and (e) within a reasonable period of time.
Developer access to your data: Our developer team has the ability to view your thunks and project-related data. They will do so only as part of investigating issues that you or other users report. By using the Thunk.AI service, you explicitly grant us permission to do so. As our service and team grows, we may amend these terms to explicitly designate certain team members who will have this level of access, and introduce further mechanisms to give you the ability to control this access.
External services: Some data associated with your thunk projects may be exposed to other services (subprocessors) utilized by Thunk.AI (including the services highlighted below) and will be governed by the data access and retention policies of those services.
OpenAI: Thunk.AI uses OpenAI (the GPT-4 LLM API provided by OpenAI) to provide intelligent services. Your project-related data may be read and used by your AI agent, which in turn interfaces with OpenAI’s models. Any interaction between your AI and ChatGPT is governed by OpenAI’s terms of use. When you allow your AI to read your documents, OpenAI’s retention and privacy policy set forth here, specifically those under API Platform, apply. If you use your own OpenAI key with Thunk.AI, you can delete the OpenAI retained data at any time by logging into your OpenAPI playground.
Monitoring usage of LLMs on Thunk.AI: We use Helicone.AI (a proxy service between our system and OpenAI) to monitor the interaction your AI has with OpenAI models. Helicone.AI’s policies are described here.
Debug logs on Sentry: We use Sentry to debug issues that you may encounter. All of this information is used only to debug issues and performance and to improve the product and your experience on Thunk.AI. Sentry’s terms of service are described here.
Usage logs on Mixpanel: We use Mixpanel to record end-user activity on our site. This information is used to understand UI friction, effective usage patterns, and improve the platform. Mixpanel’s terms of service are described here.
AI Policy: Thunk.AI does not use your data to train any AI models. Thunk.AI uses OpenAI APIs governed by OpenAI’s enterprise terms of use. We encourage you to read these terms carefully, but want to highlight that “We do not train on your business data (data from ChatGPT Team, ChatGPT Enterprise, or our API Platform)”.
Third-Party Sites: The Thunk.AI service provides you with links to third party sites. Once you click on a link to a third-party site, the privacy policy of that third party site is in effect.
Communication: The Thunk.AI service uses email and other available communication channels to interact with you once you start using the Thunk.AI service. These communication channels have their own limitations with respect to message security and privacy, and to the extent your information is communicated on those channels, it is subject to those limitations.
Opting Out: If at any time you are dissatisfied with the Thunk.AI service and want to stop being a user (we hope not!), please send an email to support@thunkai.com and we will delete your account. We may retain some information for a reasonable period for purposes of reporting and auditing. We may retain and utilize your historical non-personally identifiable usage information indefinitely as part of the data used for analysis and learning.
California Users: Under California Civil Code Section 1789.3, California users are entitled to the following consumer rights notice: If you have a question or complaint regarding the site, please send an e-mail to support@thunkai.com. You may also contact us by writing to ThunkAI Inc, 2805 170th Ave SE, Bellevue WA 98008. California residents may reach the Complaint Assistance Unit of the Division of Consumer Services of the California Department of Consumer Affairs by mail at 1625 North Market Blvd., Suite N 112, Sacramento, CA 95834, or by telephone at (916) 445-1254 or (800) 952-5210.