AI Assistant

How does the AI system work?

Our AI Assistant uses advanced language models combined with our own proprietary technology to understand and respond to user queries in real time. When a user asks a question, the system analyzes the intent behind the request, searches through your product catalog and knowledge base, and generates a relevant, contextualized response.

Where does the AI retrieve data from?

The AI Assistant uses two types of data provided by you to generate responses: your product catalog and any knowledge base content uploaded through our admin panel. This data is used exclusively at inference time. So it is provided as context for the AI to generate relevant answers to user queries. No data is shared for model training. It is only processed in real time and is not retained for other purposes.

How long do you store our data?

We store your data for as long as it is needed to provide the service. Once our contract ends, we can delete all your data.

Is the data sent to servers located outside my region (EU/USA/APAC)?

No. Our infrastructure is divided into three regions: EU, USA, and APAC. Customer data always remains within the region where the store operates and is never transferred between regions. In addition, all data is hosted exclusively within our AWS infrastructure and remains in it.

Can the AI providers leverage Doofinder's customer data for other purposes?

Neither we nor our providers, use data for any other purpose than for answering your customers queries. Once our contract ends, we can delete all your data.

You also can cancel your subscription at any time, see here.

How do we protect data and what data protection measures are in place?

Your data is protected through multiple layers of security to which we maintain regular reviews of security posture to identify and address potential vulnerabilities. No client data is shared with or retained by third-party AI model providers for training or purposes beyond what is required by their data processing terms.

Has a data policy been developed that includes the privacy and protection of sensitive data types?

Yes, our data policy is located here and our DPA here.

What is your strategy and vision regarding compliance?

Our compliance strategy is built around three pillars: data minimization, transparency, and regulatory readiness. Besides operating in accordance with the GDPR, we are actively monitoring the EU AI Act's phased implementation timeline to ensure we meet each milestone as provisions come into force. We are committed to adapting our practices proactively rather than reactively as the regulatory landscape evolves.

Will there be a clearly defined human-in-the-loop or human-on-the-loop mechanism for all critical decisions made by the AI system?

The AI Assistant does not make critical decisions. It assists end users in finding products and information by generating responses based on your product catalog and knowledge base. All responses are suggestions, not autonomous actions.

On the configuration side, your team has full control over the AI Assistant through our admin panel, including the ability to manage the knowledge base, review performance metrics, and adjust settings. Additionally, the system can be paused or disabled at any time by you.

Will system logs and decision-making logic be auditable by humans?

Yes. User messages, AI-generated responses, and performance metrics, are accessible through our admin panel. This allows your team to audit the AI Assistant's behavior and outputs. On the Doofinder side, we maintain detailed internal logs for troubleshooting and quality assurance purposes.