About Lopay
Card payments for small businesses, at less than half the market rate
Lopay is reshaping card payments with a headline rate of 0.79%, helping small and medium-sized businesses keep more of every sale. Three and a half years after launching in the UK, its lean team uses AI to manage much of the operational and regulatory workload as Lopay expands across Ireland and North America, with France, Italy and Spain next.
“There are a lot of financial and regulatory obligations as a payment company, like KYC checks and verifying each merchant’s identity, ownership, and eligibility to process payments. A lot of that used to be manual, and now we use AI in Slack.”
The challenge
Scaling critical work without scaling headcount
Lopay’s headline rate depends on running the business with uncommon efficiency. Lopay must verify every new merchant, monitor changes in account activity, and complete regulatory checks on time. Across much of the payments industry, people handle this work manually, so higher transaction volumes bring higher headcount and higher costs.
AI gave Lopay a way to scale this critical work with a lean team. But answers generated in standalone AI tools stayed with the one person who wrote the prompt. That person still had to interpret the output, copy it, and share it where colleagues could act. For a small team coordinating merchant onboarding, engineering, and marketing, each extra handoff chipped away at the speed AI promised.
“Slack is a really important communication tool between people and AI.”
How Lopay works better with Slack
When AI sits in Slack, every answer provides shared context
Lopay has run on Slack since the company started in 2021, and founder and CEO Ollie Mahoney has used it since 2014. Today, almost every employee at Lopay uses Slack channels to communicate, trigger work, and track progress. AI agents, powered by Zenning (also founded by Mahoney), work alongside AI in Slack, bringing people, data, and action into the same place.
Giving everyone the context behind the numbers
Each morning, Lopay posts its headline numbers in a Slack channel, where anyone can ask what drove a change. An AI agent searches the relevant tools and considers factors such as weather and local transport disruptions. “You can ask, ‘Why did numbers increase yesterday?’” said Mahoney. “The agent might tell us, ‘A Tube strike led more people to use taxis.’”
Because the conversation happens in a channel, one person’s question gives everyone an answer they can use immediately. Colleagues gain the same context without repeating the prompt or waiting for someone to pass it along.
Proactive AI that speaks first
Lopay’s AI agents also proactively step in as needed. For example, when Mahoney asks about the next social post or checks the status of a project, the team has about 10 minutes to respond. If there’s no answer by then, an AI agent will search integrated tools, like Notion, and share its findings in the channel.
The agent also starts conversations. When an agent spots a pattern in the day’s errors, it flags the issues, investigates the likely cause, and posts its findings without waiting for a prompt. The team gains insights even when no one thinks to look for them.
“If the AI thinks there’s something useful, it can say, ’There were a bunch of errors yesterday. I’ve looked into it for you. This is the answer.’ That’s incredible. There's nowhere else that could happen, because Slack is this melting pot of information.”
Connecting AI work across teams
Lopay applies the same pattern across the business. In engineering, an AI agent creates release pull requests and posts them in Slack, giving everyone a shared view of what shipped. In marketing, employees request asset changes in a channel, and an agent edits the image, video, or audio, sharing the new version for review. When an agent detects a change in merchant activity, it alerts support, engineering, and leadership in a channel so they can coordinate their response.
Slack connects each request, the work it triggers, and the team’s feedback. “It’s a complete feedback loop in Slack,” said Mahoney.
Faster action on alerts and partner needs
Lopay’s monitoring tools post a steady stream of operational alerts in Slack, flagging inactive merchants and account issues. On-call staff rotate through a dedicated urgent channel, while PagerDuty routes each incident to the right person.
AI agents contribute fewer messages, but each one carries more context. “The AI is way less traffic, but it’s way more valuable,” said Mahoney. “It comes in and goes, ‘By the way, Ollie, I’ve noticed something here.’” Because Lopay brings alerts, AI updates, and team conversations into Slack, its agents can investigate issues using the context the company already shared. This rich historical context shapes Mahoney’s advice to AI founders: “It’s better to integrate with Slack.”
Lopay also opens shared Slack Connect channels for strategic partners and customers. The company’s support team handles in-product requests, while Slack Connect gives partners direct access to colleagues across Lopay. Engineers can answer questions where they already work, and every conversation stays visible to the people who can help.
Offering a channel during onboarding also builds trust. “When we tell customers, ‘We can open a shared Slack channel,’ it gives them confidence,” said Mahoney. “You just know you can talk to somebody.”
Applying payments-grade controls to every Slack conversation
As Lopay brings more operational work and partner collaboration into Slack, it applies the same rigour that governs its payments business. Lopay meets PCI DSS Level 1 requirements, the card industry’s highest compliance tier. Its retention policies limit message history to a defined window across internal and Slack Connect channels, so partner conversations follow the same rules as conversations inside Lopay.
“We’re PCI Level 1 compliant at Lopay, and that works nicely with Slack,” said Mahoney. “You don’t get that with any other tool.”
Lopay also gives employees clear rules for handling customer information in Slack. Customer personal data stays in the systems that protect it, while employees use Slack to reference records without copying sensitive details. Regular training helps the team apply those rules every day. If someone shares sensitive information in Slack, colleagues delete the message and flag the incident.
“Of course, Slack still does all the old-school stuff well: I can DM a colleague and ask for an update. But today, it’s also like a mini AI control centre.”
What’s next
Taking Lopay’s model of shared AI even further
Mahoney sees Lopay’s current workflows as the basis for a broader model of work, with people and AI collaborating in shared Slack channels. “The future is proactive AI: that extra ‘almost AI’ colleague in Slack helping you with things,” he said.
As Lopay grows, Slack will keep its people, AI agents, and shared knowledge connected, giving its lean team more capacity to build, support customers, and expand.
