Listening labs
Bring a real customer conversation. Pull it apart together. Learn to hear the patterns nobody else is hearing yet.
A small, invite-only group of CX leaders comparing notes on what's actually working with AI — by dinner, by working session, by pitch night, and on WhatsApp in between.
We're also launching a venture studio — a small batch of hand-rolled conversational AI projects for the teams whose agents need to actually work in production.
Invitation only · Operator-led · Founding cohort 2026, starting in NYC











Most decisions about customers get made far away from any customer. We start by paying attention to what's actually being said.
The companies losing customers fastest don't realize it. The frustration never reaches the people who could do something about it.
Software is the easy part. Knowing who your customer is, what they've said before, and what they actually need — that's the work.
Chanl sponsors this community and partners are welcome at the table — but the conversation is led by the people doing the work, not by anyone trying to close a deal.
We share what failed, what surprised us, what we'd do differently. Everything else is LinkedIn.
We compare notes in whichever format fits the work. All small. All operator-led. Bring what you're wrestling with — or just come to listen.
Bring a real customer conversation. Pull it apart together. Learn to hear the patterns nobody else is hearing yet.
Work on an agent, a scorecard, or a workflow alongside people doing the same thing. Real feedback before it ever touches a customer.
Bring something live that's drifting. Six pairs of eyes on the problem. Leave with a plan, not a deck.
Builders show what they're shipping for CX. Short demos, honest feedback, no investor performance. Run alongside partners like Fabrik NYC and Bond AI.
Six seats, three hours, one question on the table. No agenda — just the conversation that doesn't fit anywhere else.
Hosted on Luma. Browse past recaps, RSVP, or subscribe to the calendar.
Why shipping an MCP server is a go-to-market move rather than an integration chore. We covered discovery, security, monetisation, and why a tool description is now your landing page. Hands-on — everyone left with a running server.
A working session on agents that hold customer context across channels — memory, tools, and what breaks once real customers are on the line.
Latency, interruption handling, and what a voice agent needs before it answers a real customer line.
Builders showing what they're shipping for CX. Short demos, honest feedback, no investor performance.
Pick any one — or all three. We meet in person for the deeper stuff and stay in touch online between gatherings.
Browse upcoming dinners, working sessions, and pitch nights on Luma. RSVP to anything that fits.
View calendar →Open chat for members and friends of the community. Share what you're working on. Ask questions. Compare notes between events.
Join on WhatsApp →Curious about an invite-only gathering? Send a short note about what you're working on. We read every one.
Introduce yourself ↓Chanl is opening a small venture studio for conversational AI — a handful of engagements per year, built alongside corporates and ambitious startups whose agents need to actually work in production.
Limited scope. Deep engagement. Terms shared on a call.
We believe AI should do more than answer questions — it should take action. Book the appointment, check the inventory, send the follow-up. That is what we build at Chanl, and it is why this room exists: the people hosting are the people shipping.
Chanl is built and operated by Elemente, Corp.
Product
Operations
Engineering
Engineering
Design
Advisory
If you lead a team that talks to customers — or you're building something that will — we'd like to know who's in the room. Tell us a bit about what you're working on and what you'd want to discuss.
By writing in you agree to receive event invitations. Unsubscribe anytime.
Customer-First AI is sponsored by Chanl, with support from a small circle of partners. Membership is independent — applicant lists aren't shared, and the table is for the people in the room.