Level AI is launching a suite of AI Workers built for the people who run CX operations: specialists for coaching, metric investigation, customer research, and more. See them demonstrated live on real scenarios.
Corinne FlanaganSr. Manager, QA, SmartsheetTurning data into action still lands entirely on people, and there is never enough time to do it well.
A supervisor managing 15 agents faces roughly 3,000 interactions per week. Manual preparation covers a small and unrepresentative slice. Coaching feedback reflects anecdotes rather than patterns.
A themes report takes two weeks to compile and three days to produce. By the time it reaches the business team that needs it, the decision window has closed.
Generic tools built for every use case get adopted by none of them. CX operations require depth and accuracy that only purpose-built workers deliver.
Each demo runs in the context of a single CX leader's Tuesday — three operational moments where the work stalls, and three workers that resolve each one.
The worker reads every interaction an agent has had, identifies patterns across all of them, and produces a session brief anchored to specific calls with timestamps. Jordan spent an hour last night on six calls. The worker covers the full month while he is on his way into the office.
The worker runs a cross-platform analysis across quality scores, CSAT, sentiment trends, and resolution outcomes simultaneously. It returns the finding with supporting call evidence in under 90 seconds. The data existed at 10:15. The meeting at 2pm now has an answer.
The worker scans the full transcript library using semantic understanding of what customers mean — reading intent across thousands of conversations simultaneously. It returns thematic clusters with direct customer quotes, frequency data, and trend lines across 30, 60, and 90 days.
Phyllis Fang and Devashish Srivastava open the session.
CX investment went into customer-facing automation. Coaches, analysts, and QA leads running the operation behind those interactions received none of it. Phyllis Fang frames the specific workflows that stall and why purpose-built workers close the gap.
One CX leader's Tuesday plays out as a narrative: three operational moments where the work stalls. Each worker runs live and resolves the exact moment it was built for. Coaching prep. Root cause investigation. Customer research. All three demonstrated on real data.
Live DemoCorinne Flanagan, Senior Manager of Training, Quality and Operations at Smartsheet, describes the specific workflows her team changed, what the output quality looked like, and what it took to move her leadership team from skeptical to aligned.
Every worker in this session runs on a platform that accepts a job description and builds a worker for it. Starting June 15, that capability is available to customers. Live attendees get early access before the public launch.
Exclusive to Live AttendeesThree strong questions from the audience, with customers on screen.
Phyllis Fang
VP of Marketing
Devashish Srivastava
Product Manager
Corinne Flanagan
Senior Manager, Training, Quality and Operations
“We used to debate based on feelings, but we could not quantify customer frustration. Now we have actual data and verbatim customer words. It is no longer a debate — it is facts. We are bringing the true voice of the customer to our business teams every single month.”
“What would have taken a team lead over an hour to collate is now done in less than a minute. That is a huge amount of time that team leads can now spend actively coaching rather than on administrative tasks.”
Making coaching and quality decisions based on reports that arrive after the decision window closes.
Building session briefs from roughly 2% of agent interactions, leaving most coachable moments and top-performer patterns unexamined.
Spending the majority of the week locating, filtering, and packaging data before drawing a single conclusion.
Manually reviewing 1–5% of interactions and unable to identify failure patterns across the full interaction population.
Every worker in this session runs on a platform that accepts a job description and configures a worker for it. If your team runs a workflow that was not covered today — your escalation path, your compliance audit, your onboarding review — that is what Build Your Own Worker is for.
Live attendees on June 11 get early access before the public launch on June 15.
Level AI AI Workers for coaching, metric investigation, and customer research — live on real data. Register free and get early access to Build Your Own Worker.
Reserve Your SeatFree to attend. Replay available to all registered attendees.
