Every day, banks and lenders make millions of decisions: who gets onboarded, which credit applications get approved, which transactions get flagged for AML review. For decades, those decisions were made by people working through slow, rules-based processes with legacy systems underneath them. Increasingly, the industry is asking whether AI agents can do that work better, faster, and at a fraction of the cost.
Taktile builds decision infrastructure for regulated financial institutions. The Berlin-born company, now in New York and London, is betting that dropping a foundation model into an existing workflow isn’t the answer.
The decision problem at the core of financial services
Financial institutions run on decisions. Whether a customer can open an account. Whether a business is creditworthy. Whether a transaction appears to be money laundering. These aren’t generic AI problems; they require domain-specific intelligence, auditability, regulatory compliance, and a coherent way to keep humans in the loop when it matters.
Taktile’s agentic decision platform addresses this set of challenges with a layered architecture — an AI Agent Manager, Decision Engine, Agentic Case Manager, Context Layer, and enterprise-grade infrastructure — to enable financial institutions to deploy autonomous agents across onboarding, underwriting, AML, fraud, and claims without compromising governance.
Beyond wrapping a model
Taktile argues that deploying AI in financial services requires more than connecting to an API from OpenAI or Anthropic. It requires domain-specific agent intelligence, business-user-controlled guardrails, human-in-the-loop escalation workflows, dedicated financial data context, and strict system governance. Foundation models provide the intelligence layer; the infrastructure around them is where financial deployments actually succeed or fail.
Taktile claims its customers across banking, lending, payments, and insurance experience measurable outcomes: 10% increases in approvals through smarter onboarding, up to 95% automation rates for SMB underwriting, more than 75% reduction in AML false positives, and significantly faster fraud detection.
Case study: One of the world’s largest insurers
The most telling data point in Taktile’s story is one of the world’s largest financial services firms. The global insurer already has a formal partnership with one of the top AI labs. And yet when it came to deploying AI agents across its business, this insurer chose Taktile as its strategic partner. The company has expanded across multiple business lines globally, with claims processing automation rates more than doubling.
The relationship with the insurer illustrates where value is accruing in the enterprise AI stack: the foundation model is necessary but not sufficient. The winning platforms likely orchestrate those models with auditability, workflow controls, regulatory compliance, and financial-services-specific knowledge baked in. That’s where institutions are placing their bets.
What it means for the industry
Taktile’s customer list includes Monzo, Mercury, Questrade, Ualá, and Kueski, among others. The platform currently powers more than 30 million weekly decisions for over 150 customers, and the company is expanding its agent library and building exclusive data products with partners including Equifax and Dun & Bradstreet.
The foundation model wars are getting most of the attention. The less visible question — who owns the infrastructure layer where those models actually get deployed in regulated industries — may prove equally consequential.
Disclosure: The author may have, or be considering, financial interests in the company mentioned in this article. This article does not constitute financial advice, investment recommendations, or an offer or solicitation to buy or sell any securities. The information presented is based on representations made by the company and publicly available sources. Readers should conduct their own due diligence and consult with a qualified financial advisor before making any investment decisions.
Join our 4dFI investment and education syndicate to learn more about emerging financial services models and the companies shaping them. Visit tearsheet.co/4dFI to join today.
For years, cross-border payments have had problems with fees, long settlement times, and lack of transparency. Unfortunately, these problems have been well understood yet largely unsolved. The systems underpinning international money movement were built in the 1970s and 1990s for batch processing and domestic connectivity. The era of real-time, global, and programmable transactions was out of sight then, but it isn’t now.
Stablecoins have long been positioned as a potential answer, but adoption stayed shallow. That is changing.
Avinash Chidambaram has spent his career at the intersection of payments infrastructure and emerging digital asset frameworks. As founder and CEO of Cybrid, he now builds the stablecoin rails that neobanks and fintechs use to move money across borders. He is also in the room with key stakeholders in Canada, like regulators and other founders, giving him a unique vantage point into the future of this space.
“We’re an operator,” Chidambaram says. “We’re actually doing stablecoin payments, so when we go through the process of getting registered with the regulators, when we talk to legislators, we’re trying to give them that perspective.”
Looking ahead, Chidambaram sees a market that has cleared a threshold.
Regulatory frameworks are no longer in their infancy. Enterprise appetite now centers strategic implementations rather than exploratory ones. And the infrastructure stack (compliance, custody, FX connectivity, accounting) has been bundled into APIs that product teams can actually consume.
Even as recently as 2023, Chidambaram said conversations with legislators about stablecoins were largely educational. Fast-forward to 2026, and a lot has changed. Today, they are around operations and implementation. He added that policymakers across jurisdictions have moved past the basics and are now debating finer questions: which blockchains to support, how to think about validator risks from a sovereignty standpoint, and whether a domestic sovereign stablecoin makes more sense than a central bank digital currency, he shares.
The GENIUS Act in the US is a particularly meaningful inflection point that has set a benchmark for stablecoin legislation that other regional jurisdictions are now working from. In Europe, MiCA has done similar work, and the impact is visible: roughly 200 stablecoins are now issued globally.
“[With increasing regulatory clarity,] people understand what the risks are,” Chidambaram says. “People understand what kind of capital controls are required, and now they’re really starting to understand what the value props are.”
Yet discussions about implementation don’t mean everyone is rushing to push something out the door. Now, the uphill battle is not lack of awareness but the pace of work.
National frameworks, particularly around a sovereign stablecoin issued by a central bank, take considerable time to design and implement. In the meantime, the private market moves at its own speed. Chidambaram’s read on where most legislators have landed is that they understand the value proposition and are focused on getting the capital adequacy and redemption risk questions right, rather than blocking adoption.
The structural problem with legacy rails
Legacy rails were designed for domestic batch processing, and the workarounds built on top of them have multiplied over decades into layers of intermediaries, correspondent banks, card networks, clearing houses, and separate compliance vendors. Each layer adds cost, settlement delay, and opacity. And, most importantly, creates a structural problem rather than merely a technological one.
“Modern businesses are forced to operate on top of these really fragmented systems,” Chidambaram says. “They were never designed for real-time, global, programmable payments.”
The practical consequence for a fintech or enterprise trying to automate payments within a software workflow will give any decision maker pause: Connecting an ERP’s purchase-approval logic directly to a payment execution layer means threading through all of that fragmentation, and doing so in a way that still satisfies a compliance team running AML screening, KYC, and reporting across multiple jurisdictions.
By comparison, stablecoins might almost feel like magic. Rather than trying to connect fragmented systems, stablecoins bypass intermediaries by connecting two parties on a shared ledger. While the compliance obligations remain, the technology allows those checks to run before the payment executes, rather than managing them across a relay of institutions that each see only part of the transaction. And because the transaction is on-chain, both sender and recipient know when the money has landed.
Stablecoins in production
When Cybrid walks enterprises through the stablecoin payment flow, one detail lands particularly hard: traceability.
“We talk to enterprises,” Chidambaram says, “and they’ll tell us: ‘I sent $20,000 to my supplier, and he’s like, you didn’t send me all the money.’ And everyone just agrees that this is just how it works. That’s kind of crazy.”
With stablecoin infrastructure underneath, that conversation does not happen. A company acquires US dollar-backed stablecoins, sends the payment directly to the recipient’s wallet, and can confirm receipt on-chain in roughly two minutes. The FX price is guaranteed at the time of the transaction, not settled at whatever rate applies three days later when the funds actually arrive. For enterprises buying goods from overseas suppliers, that certainty has real value on both sides of the transaction.
Once payment execution is fast and programmable, the logic can move into the software stack. An ERP that flags a low-inventory condition can trigger a supplier payment approval workflow, route the CFO a notification, and execute the transaction on approval, without anyone logging into a separate banking portal.
“That feels kind of magical, but that’s actually here today,” Chidambaram says. “And that’s what we’re enabling.”
What actually unlocked adoption
Stablecoins have existed for roughly a decade, and for most of that time they were closely associated with crypto trading infrastructure. Initially, their core value prop was a faster way to move in and out of cryptocurrencies on exchanges than routing through a bank. That association was accurate as a description of where the volume lived. But the problem is it also shaped how enterprise and institutional audiences thought about the technology. For better (and worse), they thought of stablecoins as a crypto-adjacent instrument rather than a payments tool.
With increased regulatory clarity and the maturation of the operational stack around stablecoins, the “stablecoins as crypto-adjacent” narrative has changed.
“Every fintech, every bank now has someone in their organization whose job it is to figure out what is our stablecoin strategy,” Chidambaram says.
Compliance, mixed custody, fiat connectivity, accounting that meets IFRS standards has perhaps been the most significant change, and organizations like Cybrid assembled those components and wrapped them in an API. For fintechs fielding c-suite pressure to develop a stablecoin strategy, the availability of a bundled infrastructure layer that already carries SOC 2 compliance and regulatory registrations has materially lowered the barrier to moving forward.
“You’re not thinking about, ‘How do I do this integration to all these different things?’ It’s simple. It’s an API,” says Chidambaram.
Moving from curious to live
For executives who are past the evaluation stage and ready to move toward a pilot, Chidambaram’s advice is to lead with use case specificity rather than treating stablecoins as a general infrastructure upgrade.
He sees the clearest value today in the following categories:
Cross-border supplier payments
Global contractor payments
Marketplace payouts to large and distributed recipient pools
Treasury liquidity management across multi-jurisdiction entities
Treasury liquidity management, in particular, deserves attention. Stablecoins allow a treasury function to consolidate management across subsidiaries from a single compliance team, a structural efficiency that becomes more valuable as organizations scale internationally.
“CFOs need to start to look for a service provider who offers this compliance stack with the appropriate certifications,” he says, “and then say, ‘Where do I get the most value from this?’”
The scope of integrations required to assemble stablecoin payment infrastructure independently (rails, compliance, FX, accounting, custody, regulatory registration) is substantial, and maintaining those integrations over time is an ongoing cost.
For a company whose core business is e-commerce, marketplace operations, or financial services product development, that integration burden competes with resources better spent on the actual product. The more defensible path is finding a provider that has already assembled the stack and can deliver it as a consumable layer, freeing the internal team to focus on what they are actually building.
Nine months into her role as Chief Product and Technology Officer at Temenos, Barb Morgan is focused on a simple principle when it comes to product strategy: quality over quantity. “We want to build less, but build it better,” Morgan said during a conversation at the Temenos Regional Forum Americas 2025 held May 28-30 in Miami.
Temenos’ approach centers on co-creating meaningful solutions with bank customers rather than rushing to market with multiple products. Morgan emphasized that the company is “really focused on making sure that whatever we put out there is meaningful,” as the industry navigates what she calls the “AI hype curve.”
Morgan’s insights reveal why many banks struggle with AI adoption despite the technology’s promise. The real barriers aren’t about computing power or algorithms — they’re messier problems involving decades-old data systems that were never designed for AI and organizational cultures that haven’t caught up to the pace of technological change.
Her conversation also detailed Temenos’ bet on bringing innovation closer to customers, such as through its new hub in Orlando designed for co-creation, and why the company is taking a strategic and deeply integrated approach to AI that enables banks to deploy AI-powered solutions faster and safer.
Temenos has structured its AI approach around three core components: Gen AI embedded directly into its platform and products, agentic AI with a first solution for sanctions screening already live at one Tier-1 bank, and an AI studio for custom use cases. “We have a lot of customers coming to us with very unique use cases, and so we want to provide them a platform that’s pre-built with banking modules,” Morgan explained.
The company’s focus on embedded AI addresses a common industry challenge. “Having it embedded, versus our customers trying to figure out how to bolt it onto our product, is really important to us,” she said. This approach allows banks to access AI capabilities without investing too many resources into integration.
Banks are ready for AI – their data isn’t
One of the biggest obstacles to AI adoption isn’t fear of technology, but foundational data issues, shares Morgan. “A lot of banks over the past 10 to 15 years went through this huge digital transformation, but what they didn’t transform was the data in the back end,” Morgan noted. “In order to leverage the power of AI, you have to have your data clean.”
This reality has shifted many of Temenos’ client conversations toward data readiness rather than AI capabilities. “Our clients also want to leverage their own data systems. So how clean is your data? Is it really ready? Because for secure AI products, you have to have your data in order,” she said.
Cultural change is the harder challenge
Beyond technical hurdles, banks face significant organizational resistance to AI implementation. “I was talking with one of our US banks last week, and he said I underestimated the amount of cultural change that’s necessary, because so many people are afraid of AI,” Morgan shared.
The fear stems from job displacement concerns rather than technological limitations. “They think it’s going to take my job away, versus thinking of it as augmenting their job and being more of a side by side partner,” she explained. This cultural aspect has to become a major focus for banks that want to succeed with their AI implementations.
A gradual approach to AI deployment
Temenos’ strategy acknowledges these cultural and technical challenges by allowing banks to phase in AI adoption. Morgan described how one tier-one bank using their agentic AI product FCM AI Agent started with just 5% of its traffic, then gradually increased it to 20%. “It wasn’t because they didn’t trust the technology. It was because they were getting the rest of the organization comfortable,” she said.
This incremental approach extends to customer-facing applications as well. “A lot of people, it seems, have their favorite [Gen AI] tool on their phone,” Morgan observed. “I think maybe the banks have underestimated that the customers are actually ready to interact with AI.”
Bringing innovation closer to customers
Part of Temenos’ US expansion includes the opening of its Orlando Innovation Hub, designed specifically for co-creation with bank customers. “Instead of just expanding one of our existing offices, we’re actually going into a brand new building,” Morgan said. “It’s all about being able to do the design workshop, but then the space can transform to doing co-development together.”
The facility will include spaces that can replicate bank branch environments. “There’s a space where we can make it feel like you’re walking into the branch of the bank, and so we can actually recreate exactly what it’ll feel like for their customers,” she explained.
Market-centric over centralized delivery
The Orlando hub represents a broader shift in Temenos’ delivery model. “Over the past 30 years, we have had a pretty centralized delivery team, and this is about bringing it closer to our customers,” Morgan said. “Versus centralized delivery, it’s more about market-centric innovation.”
This approach is driving the firm’s hiring, with plans underway to recruit for 200 positions at its Orland Innovation Hub. “At a recent hiring event, every candidate who received an offer accepted,” Morgan noted. “They were really excited about the co-innovation and the ability to actually work how we want and bring our best selves.”
Building products that actually ship
Morgan has instituted a new discipline around product announcements, moving away from proof-of-concepts toward deliverable solutions. “We’re only going to announce things when they’re live and ready to use now,” she said.
The company has also allocated 25% of its development capacity specifically to customer-driven features. “We’ve actually allocated about 25% of our capacity to just listening to customers and putting their needs on top of what we would already have planned,” Morgan explained.
This customer-centric approach extends to the broader organizational transformation Morgan is leading.
