Introduction
Elaras Ask is a platform for building and deploying AI agents on your website or product. You give an agent a personality, some knowledge about your business, and a set of skills — then embed it anywhere with a single script tag or API call.
What you can build
- A support agent that answers questions from your documentation and looks up order status in real time
- A lead capture agent that qualifies visitors and routes them to the right person
- An onboarding agent that walks new users through a guided setup
- A concierge that remembers returning visitors and personalises every conversation
Key concepts
Agent — the deployed AI assistant your visitors talk to. Each agent has a unique public key (cbt_...) used to embed it and call the API. An agent has a name, appearance settings, an assigned personality, a knowledge base, and a set of skills.
Personality — the configuration that shapes how your agent thinks and speaks. It contains a system prompt, a greeting message, a fallback message, and a choice of underlying AI model. Personalities are versioned — you edit a draft, test it, then publish it when it is ready.
Knowledge — the information your agent draws on when answering questions. You can add individual pages, crawl a whole website, or write precise question-and-answer pairs for facts that must never be paraphrased.
Skills — actions the agent can take during a conversation. An instant skill fires a webhook to your server with extracted parameters and uses the response in its reply. A flow is a multi-step skill that guides the user through a structured process (booking, intake, qualification) before acting.
Testing — before you publish a personality, you can test it in the preview panel and run saved scenarios. A scenario is a test case: a message (or short conversation) and the outcome you expect. The platform judges whether the agent passes.
Sessions and leads — every conversation is a session. If the agent captures a visitor's contact details, they appear as a lead you can export or push to your CRM.
Who it is for
Elaras Ask is for product teams, customer success teams, and developers who want a capable AI agent without building the infrastructure from scratch. The dashboard is designed so non-technical users can configure and iterate on agents; the developer API and webhooks are there when you need to go deeper.
What to read next
- Your first agent — step-by-step walkthrough of the setup wizard
- Agents — managing your agents in the dashboard
- Knowledge — adding content your agent can draw on
