Omie

How Omie’s support team resolves 8 in 10 conversations with Fin
Omie is a Brazilian cloud ERP platform built for small and medium-sized businesses. For many customers it sits at the center of their operations, running essential business processes from issuing invoices to financial services like payments and bank integrations. The platform also includes an education area, with free courses to support entrepreneurs and their teams as they grow. When a platform sits that close to a company’s finances, support questions are rarely trivial.
As Omie grew to serve more than 180,000 businesses, that complexity carried straight into support. Across the company, the team came to handle roughly 1.37 million conversations a year, spanning everything from invoice configuration and financial routines to tax compliance. On top of the sheer volume, the operation was made up of several specialized teams, each with its own indicators, processes, and metrics, which made tracking results and standardizing the customer experience a constant challenge. Given that volume and the company’s fast growth, the obvious answer of hiring more agents began to show its limits: the time needed to recruit, train, and ramp new teams became a critical constraint on scaling. Omie needed a new operating model, one that could absorb the company’s growth without compromising quality, efficiency, or the customer experience.
Omie wanted an AI solution it could put on the front line of every conversation. The bar was high: answers had to be accurate enough to trust with tax and financial questions, and Omie’s own team needed to own and tune the system day to day rather than wait on a vendor for every change. After weighing their options, they decided to implement Fin, an AI Agent designed for customer experience, and Intercom as their helpdesk.
Rather than flipping a switch, Omie rolled Fin out in stages. They started in their busiest channels, investing heavily in help content so Fin could resolve high-volume, informational questions on its own. From there, Omie connected Fin to their own systems and began automating more complex, step-by-step queries, like canceling a subscription or retrieving a customer’s monthly invoice and matching tax document. Now Fin can pull a customer’s account details mid-conversation, walk them through backing up their data before a cancellation, and even read the screenshots they send when something looks wrong, resolving the kinds of multi-step requests that used to require a human.
“When I don’t understand why Fin handled a conversation a certain way, I give Operator the conversation ID and it traces back exactly which content or workflow it used. Then I can ask it to add some guidance or build a procedure so Fin gets it right next time.”Behind that performance is a steady loop of monitoring and improvement. The team uses AI-powered insights to see which topics Fin struggles with and where the underlying content needs work, then feeds those improvements back in so its answers keep getting better. Fin Operator, an agent for customer operations, made that work far easier too, letting the team get to the bottom of any conversation in seconds. “When I don’t understand why Fin handled a conversation a certain way, I give Operator the conversation ID and it traces back exactly which content or workflow it used,” says Guilherme Faria, AI Quality & Performance Manager at Omie. “Then I can ask it to add some guidance or build a procedure so Fin gets it right next time.”
Today, Fin is involved in 94.8% of Omie’s support conversations and resolves about 80% of them with no human stepping in. That adds up to between 57,000 and 67,000 conversations resolved by Fin every month.
The impact goes beyond automation. As Fin takes on more of the front line, the human effort needed for each conversation has fallen sharply. The average number of teammate replies per conversation dropped from 7.49 to 3.25, a reduction of almost 60%.
That changes the economics of support. Fin absorbs overnight and peak-time volume around the clock, allowing Omie to grow without adding headcount in lockstep with conversation volume, the exact trap the team set out to avoid. Agents spend less time on repetitive how-to questions and more time on the complex financial and operational issues that genuinely require human expertise.
With that foundation in place, Omie is now extending Fin deeper into the business, from finance and accounting workflows to customer retention. For a company whose mission is to make running a business simpler for Brazil’s smaller companies, support has become a direct expression of that mission.


