Why Canapi is Leading Ethos' Seed Financing Round

Compounding Complexity

Whether you’re a DeepSeek fan or an OpenAI purist, you’ve probably thought more about AI/ML in the past two years than in the previous two decades combined. The story is familiar: In November 2022, ChatGPT sprung fully formed out of the head of Sam Altman, with an intelligence to rival Athena’s when she herself emerged from Zeus’ head millennia ago.

It's a compelling visual, but the reality is less mythical. Most organizations have been using machine learning or artificial intelligence in some capacity for decades. Banks, in particular, have been applying quantitative models since the 1970s—think credit scoring and early risk assessment models—and over the decades, this legacy has evolved into a complex, sprawling infrastructure of ML/AI systems.

Today, many banks deploy hundreds of models in production, with the largest maintaining libraries of several thousand live models across functions like risk management, fraud detection, credit assessment, and more. AI/ML models deliver powerful benefits but also pose risks. Unlike deterministic software, models degrade over time and even slight model drift in lending decisions or fraud prevention can have outsized consequences; regulators, wary of "black boxes," demand transparency and oversight. To manage these risks, banks have established model risk management (MRM) frameworks to ensure AI/ML models remain performant and compliant.

Historically, banks have relied on the "three lines of defense" framework: the first line develops and manages models, the second ensures validation and compliance, and the third conducts internal audits. While centralized compliance efforts help, MRM remains highly fragmented. Business units develop their own processes and adopt different technologies, often leading to incompatibility and inefficiencies. This patchwork of Excel spreadsheets, Word documents, tribal knowledge, and legacy technology results in inconsistent oversight, regulatory challenges, and an inability to scale innovation effectively.

Technology vendors have largely focused on first-line model development tools—observability software, data lineage tracking, and automated model monitoring. While these tools accelerate AI adoption, they are often purchased ad hoc, exacerbating fragmentation. Solutions aimed at second and third lines of defense frequently require complete process overhauls, making adoption painful and expensive and often ending in aborted implementations. Some banks attempt to build in-house solutions, but these can cost millions of dollars, take years to develop and ultimately end up increasing complexity. Many banks simply delay these investments - postponing modernization efforts altogether.

In recent years, regulatory scrutiny has intensified, making robust MRM frameworks essential. Without centralized systems, financial institutions are throwing more people at the problem rather than addressing underlying inefficiencies. In an era where AI capabilities evolve daily, scalable MRM infrastructure is no longer optional; it's essential.

Model Risk, Managed

We first met Jett and Mike almost two years ago when they were just starting BTV’s Mint Accelerator. Even then, their deep understanding of the problem space and conviction that existing solutions were inadequate was clear. As we followed their journey, every update was more impressive than the last and we quickly built our own conviction in their vision.

Today, that vision has translated into relationships with some of the largest banks in the world and a growing suite of high-impact features that deliver ROI from day one. Ethos is a powerful MRM platform built on a compassionate understanding of its users’ challenges.

The platform rests on two core pillars, the Ethos Workflow Engine and Governed Inventories. Whereas most existing solutions come with a predefined set of workflows, Ethos distinguishes itself by offering an adaptive orchestration layer that captures a bank’s processes as they are today. Highly configurable and API-integrated, it assigns clear responsibilities and captures real-time workflows—delivering immediate value while uncovering opportunities for streamlining and automation overtime.

Ethos’ platform is built for long-running processes that may take months and involve many stakeholders across many organizations within a bank. The solution ensures that everyone always knows the status of an existing project and what they are responsible for as well as what is coming next. All changes, actions, and versioning that occur during a process are also documented in the engine, simplifying audits and helping banks demonstrate compliance during regulatory reviews.

Many banks still store models in a patchwork of file systems and repositories, making tracking and auditing a nightmare. Ethos centralizes this inventory, providing a single source of truth for all models, third-party datasets, technology platforms, and open-source components. With robust versioning and change tracking, MRM teams can ensure model resiliency, consistency, and regulatory compliance more easily.

Perhaps most importantly, Ethos unlocks innovation. One of the biggest barriers to AI adoption in banks is updating MRM policies, which often require months of engineering and compliance work. Ethos transforms this painful process into a streamlined configuration exercise, enabling banks to adopt new models rapidly without compromising on risk management.

Partnering with Ethos

Adaptive orchestration software that captures the complexity of existing processes is—in our view—the most compelling type of software to be building today. Once a system fully understands a large organization’s workflows, it naturally paves the way for intelligent automation and custom solutions that drive organization-wide impact—benefits that point solutions rarely deliver.

By leading with Tier 1 bank MRM—one of the most complex, high-stakes workflows—Ethos has positioned itself to eventually expand into other bank functions and even adjacent verticals such insurance, legal, and pharma. While Ethos remains focused on bank MRM today, the potential for broader application is clear.

Beyond impressive ROI and a sleek demo, we’ve found customers buy Ethos partly because they love spending time with the team. In an environment where many software solutions feel commoditized the ability to connect deeply with C-suites as well as end-users is as meaningful a moat as any. That empathetic approach to product development will pay dividends as Mike, Jett, and the team deepen relationships and ship more solutions to their customers’ biggest challenges.

As a venture fund backed by over seventy financial institutions, Canapi's mission has always been to invest in companies that deliver strategic value to banks. Ethos checks every box—deep industry expertise, a product that solves a critical pain point, and a team dedicated to unlocking innovation. Initial reactions from our LPs have been overwhelmingly positive and we’re excited to expand on that momentum following this announcement.

For all these reasons, we’re fired up to announce our lead investment in Ethos’ Seed financing round, with participation from great investors and friends at Better Tomorrow Ventures and Capital One Ventures.