Here’s something most people in finance don’t talk about. The billing systems that run the back office were built for a world that doesn’t really exist anymore.
These typical toolkits, the ERPs, revenue recognition platforms and billing engines were designed with a simple assumption baked in: that companies charge their customers in more or less the same way every time. Same pricing tiers. Same contract structures. Predictable, repeatable, neat. And for a long time, that assumption mostly held up.
But it doesn’t anymore.
In a recent conversation with PYMNTS CEO Karen Webster, Apurv Bansal, CEO and Co-Founder of Zenskar, laid out the problem plainly. Today’s B2B deals are messy. Every customer wants something different. Sales teams negotiate custom terms, layer in usage-based pricing, tack on geographic adjustments. And finance teams are left trying to squeeze all of that into systems that were never designed to handle it.
“Collecting money is important. If you don’t get paid, you don’t run the business,” Bansal said, adding that the financial back office is overdue for a ground-up rebuild.
Advertisement: Scroll to Continue
His company just raised $15 million to do exactly that, building what he calls “AI-native financial architecture” designed to reinvent billing and revenue management.
Every Deal Is Different. And That’s the Problem
Think about what B2B pricing actually looks like today. A company might offer one customer a flat monthly rate, another a usage-based model, and a third some hybrid of the two with volume discounts and a custom payment schedule. Multiply that across hundreds or thousands of accounts, add in different geographies and currencies, and you’ve got a combinatorial nightmare.
On its own, that’s just the reality of selling. The real issue is that finance teams are trying to manage all of this variability with tools that expect every deal to look the same.
As Webster pointed out, billing complexity tends to get written off as a series of “edge cases,” even though entire companies have been built to address it. That framing understates how fundamental the problem has become.
“Sales is driving revenue, and rightly so… Trying to convince sales teams to change the way they structure their contracts is a lost cause,” Bansal said. Every new product, geography, or enterprise customer introduces variation—and it’s the finance teams that are left to reconcile the mess downstream.
What ‘AI-Native’ Actually Means Here
Bansal’s pitch isn’t about bolting a chatbot onto an existing billing system. It’s about starting over.
Traditional finance tools are rule-based. Finance teams define their pricing templates, invoice formats, revenue recognition rules, and then everything has to fit inside those guardrails. When it doesn’t, and increasingly, it doesn’t, someone on the finance team ends up doing it manually.
An AI-native system works differently. Instead of forcing contracts into rigid templates, it reads and interprets them. Parsing out the terms, understanding the pricing logic, and handling the downstream billing, invoicing, and revenue recognition automatically.
“All a human has to do is come in and hit okay,” Bansal said, describing a workflow where AI handles everything from contract execution through invoice generation, collections, revenue recognition, and journal entries.
“You can scale your business without having to linearly add people to your team,” he added. That’s what he means by “zero-touch finance.” A back office that largely runs itself.
The Trust Problem
Of course, “just let the AI handle your billing” is a hard sell. Finance is one of those areas where accuracy isn’t a nice-to-have. It’s existential. Get an invoice wrong, and you don’t just lose money. You lose a customer’s trust.
“Trust once broken is extremely hard to earn back,” Bansal said. “You only get one chance.”
That’s partly why Zenskar isn’t asking companies to rip and replace everything. Bansal is deliberately avoiding the classic startup move of demanding that customers overhaul their workflows.
“You cannot expect businesses to change their workflow,” he said. Instead, the platform plugs into the tools companies already use, their CRMs, ERPs, payment stacks, to minimize friction and speed up adoption.
Billing and revenue management have never been glamorous, but they’ve always been essential. What’s shifting is the recognition that the tools most companies rely on simply weren’t built for the way business works now.
For Bansal and Zenskar, the challenge now is execution. Proving that AI-driven automation can meet finance’s uncompromising standard for accuracy while scaling across industries that rarely do things the same way twice.
It’s a finished product wrapped around the bet that the finance stack is ready to be rebuilt. And that AI is finally capable enough to do it.
