Commerce Conversations

mFoundry was founded in 2003, pre-iPhone, to solve two problems: roughly 400 incompatible handsets and carriers controlling app distribution and billing.
The pivot that saved the business: landing Citi as a client convinced Drew to drop every other vertical and focus entirely on banking -- within a few years, mFoundry counted a third of the top 20 US banks as clients.
Drew's biggest career regret is moving straight into investing after selling mFoundry to FIS instead of taking time off first. His advice to his past self: take three to six months, then bet on yourself as an operator again rather than betting on other founders.
Near-term agentic AI in banking is landing in the unglamorous places -- fraud, AML, compliance, reconciliation, onboarding, and customer support -- because these are controlled environments where ROI is measurable and regulators are comfortable with augmentation over full autonomy.
Appetite is real, not just theoretical: 88% of finance leaders surveyed would allow some form of agentic AI in banking workflows, and 71% say AI-initiated connectivity now factors into which bank they choose. (flag: confirm source before publishing)
The long-term battleground is orchestration -- whoever controls permissions, identity, auditability, and governance across banking data captures the outsized value.
Reframe on AI errors: human-run banking workflows already carry 10-15% error rates, so an AI system that cuts costs in half while improving to an 8% error rate is a trade banks will make every time.
A $10B bank could plausibly run with roughly 50 people, per Drew -- the constraint isn't the technology (a core system is "a fairly straightforward ledger"), it's getting chartered and regulated.
Speed comparison: a team at Drift rebuilt a full agentic AI product in about four months, work that would have taken 18 months on a traditional SaaS timeline.
Drew's take on AI funding: conversations with veteran investors consistently land on "under-hyped," not overhyped, despite the volume of capital flowing in.
M&A dynamics are shifting -- competitive barriers to entry have collapsed, so outcomes increasingly hinge on go-to-market speed rather than product differentiation. Expect a small number of massive winners and a long tail of fast flame-outs.
Founder advice, delivered as a two-part rule: expect the best and worst job of your life, so stay even-keeled -- and never run out of money, because everything else is downstream of that.
Drew's outlook: the shift plays out over five to ten years, possibly faster, and will be decided by whoever wins the orchestration layer first.

Payments are commoditized; the value is in what happens before and after the transaction. This insight drove KeyBank's entire fintech partnership strategy.
Speed in banking is relative. Nine months from handshake to product launch was a standout result -- built on 10 years of trial and error refining processes.
Organizational structure matters. Having payments leadership report directly to the CEO gave KeyBank the empowerment and alignment needed to move quickly.
Stablecoins vs. tokenized deposits are not interchangeable terms. Conflating them is "intellectually sloppy" -- they have different structures, risk profiles, and use cases.
Near-term use cases for tokenized money: interbank and intrabank settlement, cross-border payments, and B2B transactions.
Consumer trust data: 75% would try stablecoins from a bank; under 4% comfortable with unregulated providers.
FIS-Circle partnership integrates stablecoin as another "lane in the payments highway."
Banks need to act now on tokenized deposits -- not because the technology is ready to deploy, but because the learning curve (risk frameworks, regulatory education, internal alignment) takes years.
Spend management is a real threat -- platforms like Ramp function as de facto digital banking experiences for SMBs, and the old rebate-based competition model no longer works.
Capital One/Brex is likely the largest capability-centered acquisition a bank has ever made, by a multiple of at least 5x.
Build vs. buy: Most banks should not try to own large-scale technology companies. Capability acquisitions and well-structured partnerships are the better path.
Keys to a successful fintech partnership: team alignment, holistic go-to-market planning, aligned incentives, and thoughtful servicing.
Jon's 2026 prediction: By year-end, tokenized deposits will move beyond just the largest banks (JP Morgan, Citi, BNY) and the technology will start becoming more broadly adopted across the industry.

In this episode of Commerce Conversations, Ysbrant Marcellus, partner at Commerce Ventures, interviews Tom Bianco, who leads Newline at Fifth Third Bank. The two discuss the evolution of embedded payments, the organizational challenges of building fintech infrastructure within traditional banks, and where the industry is heading next.

Here are the key takeaways:

Defining Embedded Payments: Tom clarifies the often-confusing terminology around "banking as a service," "embedded payments," and "embedded finance." He breaks it down into three segments: embedded payments (core payment capabilities placed at the point of need), embedded banking (adding deposits and fund storage), and embedded finance (adding credit or working capital). The industry's lack of consistent definitions has made it difficult for banks, fintechs, and regulators to have productive conversations about these different business models.

The Embedded Banking Evolution: The embedded space has evolved significantly from simply providing access to card schemes like Visa and MasterCard to file-based software integrations to today's API-driven platforms. What makes this moment particularly interesting is the convergence of stablecoins, new legislation like the GENIUS Act, and the emergence of new payment corridors. Tom emphasizes that it's not just about technology—institutions also need to bring risk management and regulatory oversight capabilities that come from decades of experience.

Building Different Culture Within Traditional Banks: One of Tom's key insights is around organizational design. Banks entering the embedded payments space face a branding paradox: how do you signal that you operate differently from traditional banking while maintaining institutional credibility? Tom describes building a team with a deliberately different culture—casual dress, different interaction styles—to show fintech clients this isn't a conventional banking relationship. The challenge is being different enough to win fintech clients without losing the trust signal that comes from being part of an established institution.

The Technology Stack and M&A Strategy: Tom discusses the advantage of owning a proprietary core banking platform rather than licensing one—it creates optionality to modify and extend capabilities in ways other banks cannot. He also explains how Fifth Third's acquisition of Rise Money allowed them to repurpose a consumer banking-as-a-service platform for commercial payment use cases, creating a "one plus one equals three" effect by combining legacy infrastructure with modern technology.

Organizational Design Choices: Tom reveals that about 95% of the Newline team is ring-fenced and dedicated exclusively to fintech clients, with cross-functional capabilities spanning go-to-market, engineering, product, and client success. However, he notes that some functions—like legal—don't make sense to fully separate because it limits career progression. Instead, they secure dedicated capacity within enterprise functions. He also describes a dual-track engineering model where dedicated engineers work alongside enterprise peers who help navigate internal governance and infrastructure.

Prerequisites for Success: Two critical factors enabled Newline's success: executive conviction from the CEO level that embedded banking is where the market is heading, and strategic investment that extends beyond just the embedded unit itself to supporting functions like legal, compliance, and risk. Tom emphasizes that execution discipline—hitting committed dates and milestones consistently—gives the market confidence, which matters when fintech clients are presenting sponsor bank choices to their boards and investors.

The Regulatory Clarity Paradox: Tom articulates a tension that many in the industry feel: while the U.S. financial services market is "dynamic by design, which is wonderful," institutions would benefit from short-term clarity (6-9 months out) on where regulations around stablecoins, industrial loan charters, and skinny payment accounts are heading. Without clarity, institutions must hedge across multiple scenarios rather than making focused bets.

Stablecoins as Potential Disruptor: Tom is "optimistically bullish" on stablecoins as potential "leapfrog technology" that could challenge the handful of mega-banks that currently control global payments infrastructure through correspondent banking networks. He sees stablecoins as a way to level the playing field for multi-regional banks, allowing them to serve clients internationally without navigating the traditional oligopoly. The open question is whether the right regulatory, risk, and oversight infrastructure will evolve quickly enough to let stablecoin growth scale safely.

The B2B AI Opportunity: While consumer agentic commerce (AI shopping agents) gets most of the headlines, Tom argues the real near-term opportunity is in B2B back-office automation. Fortune 500 companies employ thousands of people in accounts payable and receivable departments where AI could automate manual processes like invoice matching, payment approvals, and reconciliations. This represents "the shortest path to revenue" for AI-driven payment solutions, and institutions need to build the infrastructure that allows AI agents programmatic access to payment rails.

The Future is Agentic, Real-Time, and Global: Tom anchors his product strategy on three beliefs about where the industry is heading: more agentic than not, more real-time than not, and more global than not. He argues that the next generation of high-growth fintechs will ask, "Can you help me operate globally in real time?" Rather than incrementally improving legacy infrastructure, institutions need to build toolkits that are truly "borderless, real-time, and frictionless."

The Enterprise Payments Shift: Tom observes a fundamental shift in how banks approach enterprise payments. Historically, banks led with commercial lending and payments were a nice-to-have to capture full share of wallet. Now, payments technology is becoming the primary hook because companies are making bank decisions based on "what technology can you offer to automate my payment operations?" rather than lending capacity alone. The buyer of a payment solution is different from the buyer of a working capital solution, and banks are reorganizing around this reality.

The North Star Metric: When asked what single metric matters most over a five-year horizon, Tom's answer is monthly processing volume in the trillions. This metric reflects scale, client trust, execution consistency, and whether the platform is becoming essential infrastructure. For a multi-regional bank to target trillion-dollar monthly processing volumes represents a fundamental shift from traditional regional banking scale to competing with global payment infrastructure providers.

The conversation highlights both the immense opportunity and significant organizational challenges in building embedded banking infrastructure within traditional financial institutions, and offers a roadmap for how legacy players can compete in an increasingly fintech-driven world.

Commerce Conversations: Recent Episodes

Commerce Ventures