Understanding Dust: API vs. Dust Apps for AI Integration
Hello everyone ! I'm new to Dust and trying to learn more about the main principles and how the platform works. I've already asked quite a few questions to our dear Helpbot, but I think it's important to confirm or challenge some points with real experts 🙂! Our main need, for which we are considering using Dust, is to integrate AI features into our own product for our customers (e.g., providing one or more AI agents to collect customer information, answer their questions, generate reports based on the collected data, etc.). This would involve using Dust via API. My main question is about using assistants via API vs. using Dust Apps (since both seem possible). From what I understand:
Assistants, whether via API or through the Dust interface, are intended for internal use (e.g., rate limits tied to the number of workspace members, which prevents deployment in a customer-facing product).
Dust Apps, on the other hand, are designed specifically for customer-facing products (e.g., data injections, rate limit of 10k per app, etc.).
Can you confirm that when it comes to offering AI features to customers (potentially hundreds or even thousands per day) in our own product using an API, the right solution would be to look into Dust Apps, and not at using assistants via API? Thanks in advance for your feedback!
