Encore AI is betting that the most useful training material for enterprise AI agents is already inside a company: the conversations its employees have with customers every day.
The startup has raised $30 million in a Series A round led by Team8 to expand a platform that studies those interactions, identifies what helps deals and service processes move forward, and turns those patterns into AI voice agents.
From recommendation software to AI agents
Encore AI was founded in 2022 as Insait IO by CEO Dvir Ginzburg. Its first product focused on recommendation software for financial advisers and relationship managers.
The company has since rebranded as Encore AI and broadened the original idea into a system for analyzing customer conversations. Instead of only recommending actions to human professionals, the platform now uses those lessons to train and deploy AI agents.
Those agents can work alongside customer support and sales teams, or operate autonomously. According to the company, they can communicate with customers by voice or text, and they can also serve as assistants to employees by suggesting responses and tactics during live interactions.
The shift gives Encore AI a focused position in a crowded AI market. Its core claim is not simply that an AI agent can talk to customers. It is that the agent can be shaped by the specific conversations, habits and successful playbooks already used inside a company.
How interaction mining works
Ginzburg calls Encore AI's method interaction mining. The platform collects call recordings, emails and text messages, then connects that information with CRM systems.
From there, it breaks customer interactions into stages. The goal is to understand which parts of a conversation helped advance the process and which parts did not.
That structure matters because customer support and sales processes are rarely one-step events. Different moments can require different tactics. One employee may be especially effective at one point in a conversation, while another may perform better at a different stage.
Encore AI uses that analysis to build agents from the strongest parts of many employee playbooks. Ginzburg told TechCrunch that the agents may even reflect specific ways relationship managers communicate, including jokes, anecdotes or examples, because the system is designed around approaches that have worked in real customer conversations.
In plain terms, Encore AI is trying to make enterprise AI less generic. Rather than starting only from a broad model and a fixed script, it looks at how an organization already talks to customers and tries to turn the most effective patterns into repeatable behavior.
What companies can use it for
Encore AI says its platform can support both customer-facing automation and internal coaching. The same analysis that trains an AI agent can also show a company where its current process is weak.
That can include finding inefficiencies, friction points and key issues in support or sales workflows. If a customer conversation stalls at a certain stage, or if a certain approach consistently fails to move the process forward, the platform is designed to surface that pattern.
The company says its agents can be used in several ways:
- Communicating directly with customers by voice.
- Communicating directly with customers by text.
- Assisting employees during conversations.
- Recommending responses and tactics based on prior interactions.
- Helping companies identify gaps in support and sales processes.
This makes the product both an AI agent platform and an operational analysis tool. The customer-facing agent is one output, but the underlying conversation analysis can also inform how teams work.
Financial institutions are a major early focus
Encore AI has more than 40 enterprise customers globally, according to Ginzburg. The majority are financial institutions.
That customer base also shows up in the funding round. Planven, Lukatz and Garage participated, along with some banks and insurers. Encore AI said some of the financial institutions that invested had first used its product.
The company says its annual recurring revenue has increased more than 5x since it raised its seed round less than 18 months ago. Ginzburg declined to disclose exact revenue numbers or valuation.
The new funding will be used to expand U.S. sales operations and deploy the platform with more large financial institutions. That planned expansion fits the company's current emphasis on enterprise customers and financial-sector use cases.
The competitive challenge ahead
Encore AI is early to this market, but the source of its advantage could also attract pressure from much larger companies. Large CRM providers such as Salesforce, SAP, Zoho and HubSpot can build AI capabilities around their customers' data.
Ginzburg argues that access to data alone is not the same as using historical customer conversations as the foundation for AI agents. His view is that established vendors would need to change their implementation and technology stacks to make conversational history central to how agents are built.
That distinction is important for Encore AI's strategy. The company is not just selling automation at the edge of a CRM system. It is trying to make past interactions the operating material for future conversations.
The question now is whether that focus can remain differentiated as enterprise software companies move deeper into AI agents. For Encore AI, the Series A gives it more capital to push into large financial institutions and expand in the U.S. market while that category is still taking shape.