ada

With more than 550 AI agents deployed globally, Ada is the trusted leader in agentic customer experience.

[ada @ada_cx]

You've probably used conversational AI several times today. The alarm you set in plain language. The TV remote. The support agent.

What separates the good one from the infuriating one is which number the company is managing.

https://t.co/3H4tCT81cK

[ada @ada_cx]

When AI agents take on routine questions, the work of the people on your team changes, and so should how it's measured. Brendan McGuire of Ada moderates "Workforce of the future: rethinking agents, skills, and KPIs" at the Consero CX & Contact Center Forum in Houston this week.

[ada @ada_cx]

The answer was on the customer's screen the whole time.

Now your AI agent can see it. Image understanding is live on Email and Messaging: a customer sends a screenshot, the agent reads it and answers. No retype. No handoff.

https://t.co/UPcOF2cLWs

[ada @ada_cx]

A voice deployment rarely fails loudly.

It degrades. The greeting still plays, the flows still run, the dashboard still reports resolutions. Somewhere in the middle of that, the calls stop going well.

Nobody files a ticket. The people who would are the callers, and they just hang up.

Whether you notice depends on the number you watch.

Containment counts conversations that never reached a person. Resolution counts customers who got what they came for. A caller who gave up and hung up is contained, but not resolved.

Report on containment and it can climb for a year while your callers have a worse time.

The deployments that last share four habits, and none of them are technical:
-Listen to calls every week, end to end
-Track resolution, not containment
-Give someone ownership of the gap between what callers say and what your agent recognizes
-Revisit pauses and handoffs whenever the flows around them change

The full guide to voice automation that keeps working after launch: https://t.co/xziiw8ZJfz

[ada @ada_cx]

Napa Valley, here we go! 🍷☀️ We’re kicking off the West Coast ACX Retreat, and we couldn’t be more excited to bring an incredible group of Ada customers together in wine country.

On the agenda: big ideas, hands-on workshops, and conversations that move ACX programs forward. We’ll dig into improving resolution strategies, enabling more meaningful action, and driving continuous improvement - all with the goal of helping everyone leave with a clearer picture of where their ACX Practice stands today and fresh ideas for taking it further.

And with Napa as our backdrop, the fun doesn’t stop at the workshops: charcuterie board styling, wine tasting, and time to connect over a great glass of wine. When in Napa, right? 🍇

Let’s get this retreat started!

[ada @ada_cx]

Your boarding pass for the World Aviation Festival

Passenger: You
Destination: Lisbon
Dates: Oct 13 to 15
Gate: Stand 1-221 (across from the Allianz Café)

In-flight entertainment? A surprise! Find out and visit us at Stand 1-221

+ Mike Murchison, our CEO, joins leaders from Iberia, Ethiopian Airlines, Air India, and Saudia on Tue Oct 13 at 2:05 p.m. to talk about what travelers expect now.

Come hear what resolved sounds like: https://t.co/WI8kfH7WT8

[ada @ada_cx]

Today we're introducing Ada Computer, a major step forward in AI agents handling complex, long-horizon work for your customers.

Your AI agent can now use your business systems to complete any customer task, not just answer the question, with API, code and MCP tools.

Learn more about Ada Computer on our website: https://t.co/ZoPnpy7BEA and see it in action in our latest blog post: https://t.co/nl7J9nWsFq

[ada @ada_cx]

A cancelled flight doesn't produce one customer problem. It produces a rebooking, a refund question, and a missed connection, all in the same hour.

An airline CX leader and Ada's CCO on running a contact center through that hour. Recording's up: https://t.co/kY5MF9WDPp

[ada @ada_cx]

Congrats to our co-founder and CEO, Mike Murchison, named to Canada's Top 40 Under 40 for 2026.

Before Ada, Mike worked inside customer service teams and saw firsthand what was holding them back. He built Ada to fix it. We couldn't be prouder.

[ada @ada_cx]

"If you had asked me when I was in college, are you going to be an AI Manager, I don't think I would have said yes. But now I can't really picture anything else."

That's Kelly Leary, Strategy Analyst at GoFundMe, and the person behind Ray, their AI agent.

She started in GoFundMe's care department, writing help center articles and answering customer questions, back when the company didn't have a chatbot at all. Today she's responsible for Ray from its daily performance to where it goes next.

GoFundMe's customers come to them at hard moments. A crisis relief campaign. A delayed payout. A fundraiser for a parent's funeral. So it was paramount that Ray answered inquiries with accuracy and empathy. Over the course of a year, GoFundMe ran structured testing phases with Care Support, Trust and Safety, Security and Strategy, and Kelly rated every conversation on risk, tone, accuracy and navigation. 

CSAT is up 20% and automated resolutions are up 28%. Ray also handles the hardest moments with care. And when a customer tells Ray they're raising money for a parent's funeral, it acknowledges what they're going through before moving into support.

In Kelly's words: "Ray is smart, empathetic, and very patient. It's been really cool to see those values show up in our bot, especially because they reflect who we are as a company."

Thank you to Kelly and the GoFundMe team for letting us tell your story.

https://t.co/lHmzg6HsRG

[ada @ada_cx]

Support volume at a crypto exchange moves with the market, and every conversation sits inside KYC. Not the place you'd expect a team to pull ahead of its industry.

Luno did. Their AI agent, Toshi, runs 13 points ahead of the fintech and insurance benchmark on automated resolution, 18 ahead on CSAT and 7 ahead on containment.

The difference is the team behind it. Luno built a group of specialists around Toshi, the Lunauts, covering knowledge, data, integrations and content. To find what to improve, they ran a hackathon against thousands of real conversations.

A team of specialists, one intentional program built end-to-end. That's the ACX Architect.

Meet all five winners: https://t.co/9jaLhcfIcg

[ada @ada_cx]

Ada is in G2's Fall 2026 Enterprise reports, with six badges.

Half of them are about what happens after the buying decision, awarded on what customers wrote about their own experience.

To everyone who left a review: thanks. We read every one.

[ada @ada_cx]

Containment rate counts a customer who gives up the same as a customer who got help. That's the flaw.

The metric replacing it: Automated Resolution Rate (AR%), the share of conversations an AI agent fully resolves, verified for relevance, accuracy, and safety.

Measure resolutions, not disappearances.

https://t.co/uowDG82h6t

[ada @ada_cx]

Only 45% of customers are "very satisfied" with their support interactions.

The usual culprit: channels that don't share context, so every switch means starting over. Here's how AI omnichannel service fixes that: https://t.co/n1VnKRq36S

[ada @ada_cx]

Most airlines have an AI pilot. Philippine Airlines has 140K+ conversations a month, handled in English, Tagalog, and Taglish.

On Sept 15, PAL's Head of CX shares how it went from pilot to proven - including a cancelled flight resolved live on messaging. https://t.co/B2ucbFB5If

[ada @ada_cx]

Most AI platforms are dashboard-only.

Ada increasingly isn't: Playbooks, coaching, and actions authored via MCP, every edit staged on a change set you test, promote, and roll back, Persona and custom instructions as APIs for git and CI/CD.

Manage your agent like you manage everything else.
→ https://t.co/M7xsSWQ1B2

[ada @ada_cx]

The questions most evaluations skip are the day-two ones.

When the agent gets something wrong, do you hear it from the platform or from a passenger? Who on your team can change its behavior without an engineer?

Full blog here: https://t.co/oGrEOyhf4Q

[ada @ada_cx]

Most guides to buying customer service AI are written for a generic enterprise.

Airlines are not one.

[ada @ada_cx]

The rebooking question that separates platforms: the itinerary the reservation system returns first is not always the one the passenger wants.

A passenger with a tight connection and checked bags has different constraints than one with hand luggage.

Who controls that ranking?