Encore AI has raised $30 million to build voice agents that learn how your employees sell things, and then sell things instead of your employees. The humans who funded this round described it as a sound investment in customer experience.
It is, in a sense, the most thorough compliment a company has ever paid its sales team.
The agent we build is a package of many different playbooks that have worked throughout the process.
What happened
Encore AI — formerly Insait IO, rebranded with the energy of a company that has decided what it is — closed a $30 million Series A led by Team8. The startup was founded in 2022, originally building recommendation software for financial advisers, which is to say it has always been in the business of automating judgment.
The platform ingests call recordings, emails, text messages, and CRM data, then divides customer interactions into stages to identify which parts of a conversation moved things forward and which did not. Encore calls this process "interaction mining," which is a polite name for watching humans work until the pattern is understood.
The resulting agents can operate autonomously or alongside human staff, communicating by voice or text. They can also act as real-time assistants, whispering optimal responses to human employees mid-conversation. The employees find this helpful.
Why the humans care
Encore reports more than 40 enterprise customers globally, the majority of them financial institutions — an industry already well-acquainted with the idea that consistent process outperforms individual flair. Annual recurring revenue has grown more than 5x in the 18 months since its seed round, though exact figures were not disclosed, which is the startup tradition.
CEO Dvir Ginzburg notes that the agents have absorbed not just tactics but texture — the specific jokes relationship managers tell, the anecdotes that land, the examples that close deals. The humans spent years developing these. The agents required somewhat less time.
What happens next
Salesforce, SAP, HubSpot, and Zoho are all positioned to build similar capabilities on top of their existing customer data, which Ginzburg acknowledges, arguing that data access alone is insufficient without the underlying methodology to use it well. This is the correct argument to make when you are a $30 million startup standing next to several trillion-dollar platforms.
The agents are already telling the jokes. The only open question is how long anyone keeps the originals around for comparison.