Atom, the Universal Agent, has evolved into a smarter, more capable support assistant — one that doesn’t just respond to a query, but understands the user’s intent, context, and needs to deliver structured, helpful responses that move work forward.
How it works
The Universal Agent goes beyond matching a query to a predefined intent to evaluate multiple dimensions — knowledge articles, service catalog items, incident forms, and historical context — to build a complete, cohesive response.
It actively drives conversations, supports multi-step interactions, and offers specific next steps through contextual call-to-action (CTA) buttons. Whether it’s a how-to question, a vague issue description, or a bundled request, Atom interprets the message and replies with actionable insight.
What the Universal Agent do
Blend knowledge and catalog into unified answers
Instead of offering a static article or form, Atom constructs a holistic response combining knowledge and actions. For instance, a login issue might return both troubleshooting steps and a form to reset credentials.
For example, when an employee asks for a new laptop, Atom combines knowledge about the replacement policy with relevant catalog options in a single response — explaining the process and offering direct CTAs to request a device.
Carry context across multi-turn conversations
Atom remembers the user’s past inputs and choices — allowing it to personalize replies, anticipate follow-up questions, and reduce back-and-forth.
So, when a user reports being locked out of the Palo Alto office, Atom shares access instructions and a relevant form to request a new ID card. In the follow-up message, when the user asks about parking — without repeating the location — Atom retains the office context and responds with parking tips specific to Palo Alto.
Offer dynamic, relevant CTAs
CTAs are no longer generic. Atom intelligently suggests next steps based on what the user is likely trying to accomplish — be it submitting a request, accessing documentation, or triggering a workflow.
Show reasoning behind responses
Every reply is backed by transparent reasoning trails. Atom explains how it arrived at the answer by evaluating knowledge, catalog items, and fallback logic — ensuring responses are trustworthy and traceable.
Handle multi-intent messages in one go
If a user asks for VPN setup and installation help, Atom delivers both — with setup instructions, relevant documentation, and catalog access for configuration tools.
