Elevate support (and discover new insights)
Jump to:
What does this do?
Instantly answer users’ questions (asking them to hunt for answers is so old school). Your customers will love it and you’ll … - Learn what they struggle with - See exactly how your Docs help them (or not) - Boost retention (and lower support costs) by using these insights to optimize your product, docs and user experience
Directions
You can prep this recipe in under a minute. Just click the link below. You’ll be prompted for the path to your User Docs, an LLM API key, and your email address. Go to lunch and when you return, you’ll have an email waiting. Drop the attached widget onto your Docs page to deliver answers to users (and insights to you) in an instant. Best of all, you’ll be onboarded and ready to use Lamatic to deploy custom GenAI APIs to the edge in minutes.
How it Works?
The Docs Answerbot template includes 3 workflows.
- Data workflow: the first is a data workflow that crawls, chunks, vectorizes and indexes your User Docs (text, images and audio). By default, it runs daily to capture incremental updates (but you can easily adjust the frequency or trigger it manually). [Visual: show the Lamatic visual workflow editor with node types for this workflow]
- Retrieval Augmented Generation (RAG) workflow: triggered when a user asks a question (using the supplied widget). This edge-hosted workflow instantly vectorizes the question and performs a vector similarity search on your User Docs to provide a detailed answer (with citations) to the question. Users can explore further with follow-up questions and can also provide feedback on the helpfulness of the response. [Visual: show the Lamatic visual workflow editor with node types for this workflow] Analyst Agent workflow: is constantly analyzing user interactions looking for ways to improve your documentation and UX. It delivers its recommendations interactively via email or Slack. [Visual: show the Lamatic visual workflow editor with node types for this workflow]
Nodes Used
