How Flex Achieved a 5x ROI and Put $50M in Idle Cash to Work With Atlar’s AI Agents
“The investment in Atlar has paid for itself five times over. The obvious wins are risk reduction and seeing everything in real time. But the biggest return is the interest on cash that used to sit idle.”
THE CHALLENGE
Nine banks, two hours a day, and no central view of cash
Every morning, the same ping landed from Flex’s CEO: what’s our cash position? Answering it took until noon.
“I’d log into our various systems, pull the daily cash, compile it into a spreadsheet, then build the forecast,” says Brian, Head of Finance and Operations at Flex. “By the time I produced the report, the numbers had usually already changed. Given the volume of cash we move in a day, that was always stressful.”
Flex is one of the fastest-growing financial platforms in the US, building the AI-native private bank for business owners. The San Francisco company has raised over $100 million in equity funding, including a $60 million Series B, and grown revenue 4x year-over-year.
“We were trying to run the finance team as lean as possible,” Brian says. “Every day we spent about two hours logging into different banking portals just to understand the cash at the bank. We had seven, eight, nine bank partners and no central visibility across any of them.” Some of the bank portals were decades-old systems that refreshed balances daily at best, so a current number was often impossible to pin down.
Every payment at Flex requires at least two levels of approval, but not every bank portal supported that, so the gap was closed by hand. That kind of manual patching was never going to survive the company’s growth plans: “As we scale we’ll have 12 or 14 banking portals and hundreds of accounts across various banks, supporting business owners internationally. We needed something that could carry that.”
THE SOLUTION
One platform to see cash, move money, and enforce controls
Flex evaluated the market against three requirements: real-time bank connectivity, an enterprise-grade platform, and an interface the team would actually want to open every morning. The decision came down to a live test.
“We knew Atlar was the right choice when we made a payment through the bank portal and watched it appear in Atlar in real time,” Brian says. “And not on one account. Across all of our different banking partners.”
Behind that sits the layer Atlar spent years building before putting any AI on top of it: direct connections to banks in more than 100 countries, built and maintained in-house, carrying over $500 billion in annualized transaction volume. Execution mattered as much as visibility. “What has really differentiated Atlar from everyone else in this space is that we can send payments in Atlar as well. One place to see all of our cash accounts, manage liquidity, and move money.”
Approval chains moved into the same platform, applied consistently across every bank rather than depending on what each portal happened to support. Security ran through the whole evaluation, too; for a company managing hundreds of millions of dollars, Atlar’s ISO 27001 and SOC 2 compliance “gave us confidence they were taking the right steps to keep our information safe.”
As an AI-native company itself, Flex also wanted a treasury platform built the same way. “We were on the forefront of AI in our space in fintech, and we felt Atlar mirrored us. Legacy solutions weren’t built with an AI-first mentality. They added AI after the fact. Atlar was truly built on an AI foundation.”
The rollout, in early 2026, took under six weeks and required about half an hour a week from Flex’s side, with Atlar managing the bank onboarding from start to finish. “This is one of the easiest implementations we’ve been through,” Brian says. “There are no external consultants, everyone is in-house. Sometimes we joke that we must be Atlar’s only customer, given the level of support we get.”
THE RESULT
A 5x ROI, $50M earning yield, and mornings run by AI agents
The reporting cycle that used to stretch until noon now finishes before the working day starts.
“Every morning we wake up to a ping with the daily cash position, the cash that was spent yesterday and where it went, and our forecasted runway,” Brian says. “That’s done by 8am. The human error and the timing have been taken out of it.”
An AI agent delivers the morning briefing, and forecasting runs across 7, 14, 30, 60, and 90-day horizons on the same live data. For the two people who ran daily cash, the shift returns more than 1,000 hours a year — capacity Flex would otherwise have had to hire for. “This has allowed me more time to do things that actually provide value to the business: understanding the numbers, and figuring out how we optimize for the best outcome in any scenario we may face.”
Flex’s clearest financial return came from cash that had been sitting still. “We’ve set up AI agents to sweep cash out of our operating funds every night into overnight liquidity funds that earn interest,” Brian says. Each transfer is logged and surfaced for review. More than $50 million that previously earned nothing now earns yield, delivering a 5x return on the platform in its first year.
The same data flows into accounting through Atlar's native Campfire integration. “Campfire gave us the next-gen ERP foundation, and Atlar gave us the enterprise banking layer to manage our multiple banking partners,” Brian says. “Being able to handle billions of dollars in financial flows in real time is mission critical for our business.”
And it doesn’t stop at treasury. “One of our core focus areas this year was using as much AI and as many agents as possible across the finance team,” Brian says. “Atlar completely unlocked AI in treasury and capital markets, which means we now have time to focus on AI in accounting and budgeting as well.”
The ping still arrives at Flex each morning. It used to come from the CEO, asking for the cash position; now it comes from an Atlar AI agent, carrying the answer.
See how high-growth finance teams gain control and save time with Atlar. Reach out or request a demo.
