Product Talk

Level AI Head of Product on Building Trusted Agentic AI for Customer Experience

·38 min·3 clips
Level AI's technology detects self-harm signals in insurance calls, triggering immediate human escalation.
Ambiguity is the doorway in. Sirisha describes product leadership as making the unclear workable: giving people enough structure, getting the team pointed in the same direction, and keeping work moving. The conversation does not treat AI like a shiny side project. She talks about it as a serious priority that still needs careful testing in places where the product has not been pushed hard enough yet. Adoption comes down to small choices people can feel. Sirisha points out that technologists often miss how non-technical users actually meet a product, especially when a design detail looks minor inside the team but changes whether users trust it. The host brings the discussion into customer experience. That becomes the main example, with AI sitting inside live human interactions and then reading what happened afterward for brands. Healthcare gives the idea more weight. The host compares the problem to physician visits, where documentation can steal attention from the patient during already limited time. The workflow issue is plain: electronic health record documentation changed how conversations get captured, with structured coding now part of the job instead of later transcription or review doing all the work. Scarcity makes that harder. When there are not enough physicians, tools that organize conversational detail may take some drag out of the process. The example turns emotional clues into product requirements. Depression, post-traumatic stress, and other concerns can show up in conversation even when the clinician does not document them clearly. Analytics become more than reporting here. If the system can read structured content after the visit, it can help catch sentiment and missed signals that matter. The interview closes warmly, with Sirisha saying the questions got her thinking and the host placing the series around future AI leaders from any industry.

As heard by us

A grounded product-minded look at agentic AI in customer conversations.

This conversation with Sirisha, head of product and design at Level AI, frames agentic AI as a product problem as much as a technical one. It is most useful when it stays with the work of turning ambiguity into structure: aligning teams, making complex systems legible, and…

Read the full review in PlayNext →

Why you'd press play

You will hear AI treated as a product discipline: turning messy conversations into structure people can actually use.

Read the full recommendation in PlayNext →
Listen to the show on