What are Agents?

Agents are AI team members that can work alone, or as part of a team of other agents and real people. Agents have clearly defined roles and responsibilities, and they can choose how best to get their work done using the domain-specific knowledge you teach them, and the tools you equip them with.

Get started with agents

There are a few ways you can get started with building your own agents:

Your AI Agent Workforce


Scales to meet demand

You can scale agents and multi-agent teams up and down to meet demand. Many of our customers experience seasonal fluctuations in demand, with some time periods or regions requiring more support than others (E.g. Education, hospitality and finance industries).


Increases engagement

You can use agents to automate repetitive time-consuming tasks, so that your human employees can be freed up to do more engaging work. Our sales customers love that they can spend more time on customer meetings and closing deals instead of spending so much time qualifying leads. Their BDR agent frees let them move most of there efforts further down the funnel.


Learns from experience

If your agent cannot answer a question it has been asked by a customer, it can loop in a sales rep or a manager via your preferred communications tool (email, instant chat etc), and store the answer for future use. We send our “I don’t know” escalation into a Slack channel. Anyone in our team can just click on the link in there, come into Relevance and give it an answer. Our agent will then rephrase that and answer our customer. You can also insert it into an FAQ knowledge base if you want.


Can work autonomously or on co-pilot

Agents on our platform can complete work end-to-end without you needing to be involved at all. For example, we have an agent that checks for new public tools that have been released on our platform every weekday at 9am, and then produces tool explainer webpages and uploads them to Webflow without needing our input at all. The best use-cases for autopilot agents are those that are low-risk (nothing terrible will happen if something goes wrong).


Adaptive and non-deterministic

Agents do not have to respond to tasks in the exact same way ever time like traditional software solutions. They have the ability to choose the most appropriate way to respond on a case-by-case basis. E.g. A customer support agent might be asked to help a customer resolve a billing issue on one day, or to troubleshoot a technical bug on another day.