Scaling smarter: How Gamma supports 50M users with 20 agents and Fin

Gamma.app helps people create stunning, AI-powered presentations and websites in seconds. But as their user base soared past 50 million, their small support team struggled to keep up. Inbound conversation volumes regularly hit 30,000 per month, with requests coming from across the globe in multiple languages and across time zones.

Gamma needed a new model for scale. “We’ve always believed in scaling leanly – not just hiring to throw bodies at problems, but using the solutions we already have at our fingertips,” says Hilary Dudek, Head of Customer Experience. Guided by that mindset, Gamma turned to Intercom’s AI Agent, Fin, to help handle the volume, and ultimately become a central part of how support works: always-on, multilingual, and smart enough to act.

Since its implementation, Fin is now involved in 100% of inbound support conversations and resolves 72% of them end-to-end. It delivers over 18,000 resolutions each month, in multiple languages, around the clock – reducing manual handling from 94% to just 24%, and keeping CSAT steady at 84%.

Beyond scaling the support Gamma’s small team can provide, Fin has created space for them to focus on more strategic, meaningful work that drives the business forward.

Here’s how.

As Gamma’s product gained traction globally, demand for support surged. Conversations were coming in around the clock, across multiple time zones, and in multiple languages – but the lean CX team wasn’t built to scale at the same pace.

They faced three clear challenges:

• Rapidly rising volume: Monthly inbound conversations regularly exceeded 20,000, putting consistent pressure on the queue and their team.
• Limited human coverage: With a small in-house team and outsourced agents, they couldn’t be available 24/7, or staff for every region or language.
• Maintaining high-quality responses: Customers expected fast, accurate, and personalized support, regardless of whether they were free or paid users.

“We were getting hammered with incoming conversations,” says Hilary. The team’s median first-response time had climbed to 90 minutes, and their CSAT had dropped to around 69%. “Wait times were long, customers were getting frustrated, and our support agents were burning out,” Hilary shares. “So we needed a way to provide instant, high-quality responses without doubling or tripling our team.”

But they didn’t just want to automate. Instead, Gamma was set on building a support system that could match the ambition and pace of their product.

Gamma didn’t rush into AI support. When they first trialed Fin in early 2024, it didn’t deliver the results they were hoping for. “Our volume was still manageable at the time, so it was easy to postpone,” remembers Hilary.

But by the end of the year, the landscape had shifted. Volume was climbing, wait times were growing, and the team was feeling the strain. With the holidays approaching and limited coverage available, Gamma needed a solution that could step in fast – and work reliably while most of the team was offline. Hilary began exploring alternatives to Fin and even started migrating data to another vendor, but hit roadblocks around support and transparency.

Meanwhile, Intercom had evolved too. “They really stepped up as a partner,” Hilary says. “They addressed things we’d flagged, put us into key betas, and made it feel like they were in this with us.”

When the holidays rolled around and most of the team were offline, Hilary made the call to switch Fin back on. And this time, it stuck. “Gamma has a culture of experimenting quickly,” Hilary explains. “We’re big believers in failing fast – so we decided to give Fin another shot, see what Fin could do during that holiday period. What did we have to lose? We had everything to gain.”

My perception of what Fin could do completely changed. It went from being just another bot to something we could trust – smart enough to learn fast, pick up nuance, and deliver accurate answers.

With no internal coverage and only limited BPO availability, Fin handled 100% of inbound support conversations. “We set Fin live and let it run,” Hilary shares. “It was handling every conversation and doing incredibly well! Right upfront, it was resolving 50% of conversations, deflecting 72%, and holding a 75% CSAT. For a first rollout, that was huge.”

It exceeded expectations at one of the busiest periods of the year – detecting sentiment, routing smartly, and even flagging edge cases the team hadn’t anticipated. “My perception of what Fin could do completely changed,” Hilary reflects. “It went from being just another bot to something we could trust – smart enough to learn fast, pick up nuance, and deliver accurate answers.” 

In one standout example, Fin escalated a case when it saw a user struggling to pay – even though their metadata marked them as free. Subtle signals like these made a big impact. “It more than did its job,” Hilary remarks. “That was definitely a turning point for me. Having full visibility into how Fin made decisions – being able to inspect the logic and retrace its steps – that’s what gave us real confidence.”

Today, Gamma’s support model looks nothing like it did a year ago. Manual handling has dropped from 94% to 24%, even as conversation volume continues to grow. Fin now resolves 72% of all conversations, engages in nearly every single one, and delivers 24/7 global support coverage with a steady 84% CSAT.

This has freed up the lean human team – now comprising just 20 outsourced agents, two internal technical support engineers, and a single community manager – to focus on higher-value work while supporting 50 million users. “The team breathed a sigh of relief,” Hilary shares. “Fin handles so much that they finally have room to focus on interesting, challenging work, explore our product, and deliver a more thoughtful customer experience.”

Fin is more than a deflection layer – it’s deeply embedded into how Gamma’s support function operates. Free users are routed directly to Fin, while custom escalation workflows ensure that Fin hands off only when appropriate – routing conversations based on user tier, payment signals, and sentiment. Fin draws from Gamma’s help center and internal knowledge to deliver accurate, brand-aligned answers. It automatically translates conversations in real time, helping customers in their preferred language without the need for multilingual agents.

Custom workflows help Fin triage and close out cold outreach emails, detect billing issues, and handle refund requests and payment failures through Stripe integrations. 

“Fin doesn’t just respond,” Hilary remarks. “It acts. It adapts. And it does it at a scale we never could have imagined.”

Building on that success, Gamma began introducing Fin across their broader ecosystem. “Fin now helps in our Slack community, answering advanced product questions from our global users in the middle of the night,” Hilary shares. “We’ve also fed Fin our API documentation so it can start supporting developers using the beta and triage bugs directly to engineers.”

Gamma’s next chapter is about deepening automation and designing support for the long term. The team is expanding Fin’s role across their ecosystem by automating tasks like refunds, bug logging, and feature requests, evolving proactive workflows, and embedding AI deeper into community support.

Fin isn’t replacing humans – it’s making the work more meaningful, more strategic, more sustainable.

They’re also rethinking roles. New responsibilities are emerging around workflow strategy, AI training, and content optimization – all areas that didn’t exist a year ago. “I gave a talk recently on scaling CX leanly,” Hilary says. “And this is it. Fin isn’t replacing humans – it’s making the work more meaningful, more strategic, more sustainable.”

For Gamma, support is no longer just keeping up with growth. It’s enabling it.