Why Capital One built its multi-agent AI platform around open-weight models
Our take
Capital One’s recent unveiling of its multi-agent AI platform, built around deeply customized open-weight models, signals a significant shift in how financial institutions approach AI adoption. As discussed at VB Transform 2026, the bank isn't simply integrating existing AI tools; they are actively *building* AI tailored to their specific needs. This strategy contrasts sharply with the prevalent reliance on off-the-shelf foundation models, a point underscored by exploring New EU Guidelines For AI Labelling which highlights the evolving regulatory landscape demanding greater transparency and accountability in AI systems. The decision to prioritize customization, leveraging Capital One's proprietary data, is a bold move that positions them to gain a distinct competitive advantage, particularly in a sector where data security and regulatory compliance are paramount. They are essentially recognizing that generic AI, while powerful, lacks the nuance and contextual understanding necessary to truly excel in complex financial operations.
The architecture itself – a centralized, enterprise-wide platform with built-in governance and a multi-agent orchestration harness – demonstrates a commitment to scalability and control. The MACAW workflow, employed in fraud detection, is a compelling illustration of this approach. Routing interactions through specialized agents, each with specific roles like understanding, reasoning, validation, and explanation, allows for a level of precision and accuracy difficult to achieve with a single, monolithic model. This echoes a broader trend in AI development, moving away from "one-size-fits-all" solutions toward modular, adaptable systems. The benefits extend beyond the immediate use case, as Vanee notes, with customizations yielding positive effects across the entire portfolio. This “general lift” highlights the potential for compounding returns on investment in custom AI development, a concept further explored in Constraining Output Space for SLM Narrow Automation Optimization, which focuses on optimizing narrow automation tasks within SLMs.
Capital One’s embrace of open-weight models is particularly noteworthy. While these models offer flexibility and transparency compared to closed-source alternatives, they also demand significant internal expertise for customization and maintenance. However, the rewards appear to be substantial, allowing the bank to harness the power of large language models while retaining control over their data and ensuring alignment with their specific business requirements. The development of an agentic research system for optimizing backend infrastructure further exemplifies this proactive approach, enabling researchers to efficiently explore and implement performance improvements. The ability to automate this process, as described by Vanee, reduces human error and accelerates innovation, representing a powerful advantage in a rapidly evolving technological landscape. Looking at the challenges faced in mortgage lending, as outlined in [Looking for real-world examples of predictive analytics in mortgage lending [D]](https://venturebeat.com/post/looking-for-real-world-examples-of-predictive-analytics-in-m-cmsrgzg6a0cvfmi9zb09ggv4t), it’s clear that customized AI solutions are increasingly critical for achieving measurable business outcomes.
The forward-looking insights shared by Vanee – the rise of model routing for improved accuracy and the shift towards proactive, event-driven AI – paint a compelling picture of the future of enterprise AI. The ability for AI systems to anticipate and respond to events without explicit human prompts holds immense potential for enhancing efficiency, mitigating risk, and delivering more personalized customer experiences. However, the deployment of such proactive agents will undoubtedly require rigorous testing and monitoring to ensure safety and reliability. Ultimately, Capital One’s journey demonstrates that true AI leadership isn't about adopting the latest buzzword, but about building a robust, adaptable, and deeply integrated AI platform that aligns with specific business needs and unlocks transformative value. The question now is, will other financial institutions follow suit, or will they remain reliant on generic, off-the-shelf solutions?
Presented by Capital One
At VB Transform 2026, Kel Vanee, MVP of machine learning engineering at Capital One, spoke with Sam Witteveen, Senior Technology Contributor at VentureBeat, about how the bank built a scalable multi-agent AI architecture around deeply customized open-weight models rather than relying on an off-the-shelf foundation model.
"At Capital One, we're not just using AI, we're building AI," Vanee said.
The groundwork was laid years ago with Capital One's early investments in data transformation and cloud adoption, which Vanee said were foundational to moving quickly when the current wave of AI arrived. That technical foundation enabled the company to make several deliberate architectural decisions, including building a centralized, enterprise-wide AI platform with built-in governance, deeply customizing open models with proprietary data, and constructing its own multi-agent orchestration harness.
Customizing open-weight models with proprietary data
Rather than relying solely on off-the-shelf frontier models, Capital One fine-tunes open-weight models using its rich, proprietary data.
"We view our data as a huge advantage and something that nobody else has, something that the general frontier models cannot provide. So we are taking that data and deeply customizing these models," Vanee explained. He added that real-time data is absolutely critical to bring in fresh context during live customer or associate interactions.
Vanee also revealed an unexpected benefit of this approach: extensibility across the enterprise.
“As we customize those open-source models for one use case, we actually see benefits across our whole portfolio," he noted. "We are training that model to be an expert at Capital One use cases, policy, and nomenclature. As we do that training, we see a general lift."
Inside Capital One's multi-agentic AI workflow
As an example of the approach, Vanee pointed to a customer-service workflow for bank fraud that handles millions of calls a year, where interactions range from roughly four minutes to as long as sixty minutes, and where an initial attempt at engaging a single large language model proved insufficient. With Capital One's multi-agentic workflow (MACAW), interactions are routed through specialized agents with governance and guardrails built in.
"The MACAW workflow is made up of a number of different agents," he said. "The first one is an understanding agent. Its purpose is to look at what the customer is saying and try to understand what their intention is.”
From there, a reasoning agent is given several specific instructions to generate a summary; a validation agent fact-checks the summary to ensure it is accurate; and an explaining agent turns the summary into a formatted document with all necessary details that is then shared with agents.
For the consumer banking use case, this workflow helps several hundred customer-service agents who specialize in complex fraud calls. The post-call summaries it generates help document long, back-and-forth interactions that agents previously had to reconstruct by hand.
Capital One’s multi-agentic architecture also underpins Chat Concierge, a customer-facing auto-shopping assistant, which further leverages a version of Meta's open-weight Llama model that has been customized with Capital One's proprietary data. It uses the same division of labor, with one agent conversing with the customer, one building an action plan from business rules, one evaluating accuracy, and one explaining and validating the result.
Optimizing latency and cost with an agentic research system
Beyond customer-facing solutions, Capital One is also leveraging agentic AI to automate rote tasks for its employees and help them focus on high-leverage aspects of their work. In one example, the company built an autonomous agentic optimization solution to tune backend hosting infrastructure.
Vanee explained that in the world of LLMs, where new optimizations are delivered every day, they aren't all complementary. Combining two good optimizations can sometimes cause a performance regression.
"This agentic system will run through a search space that is designed by the researcher, handle all the mechanics of setting up that experiment and running the experiment, and then put a whole summarization of the results in front of the researcher," Vanee said.
Vanee added that the system allows researchers to “find the series of optimizations and configurations that's really going to give [them] the best latency possible.”
What's next: model routing and proactive, event-driven AI
Looking ahead, one big trend Vanee sees is routing abstraction layers that a platform seeks to validate over multiple models, both for cost and accuracy.
"We actually think that you can get better accuracy than any individual model simply by routing across a broader set of available models, because different models are going to excel in different areas," he said.
His second prediction was a shift toward systems that act without waiting to be asked, while also emphasizing that deploying such proactive agents would demand rigorous testing and monitoring.
"The thing I think is going to become bigger in the future is more proactive and event-driven AI," Vanee said. Rather than waiting for a human prompt, AI would step in as soon as it detects conditions that warrant action.
"This is going to enable more monitoring and larger-scale monitoring, and it'll empower us as we fight fraud and address these opportunities," Vanee said. "So proactive AI is going to be a really important trend."
Driving continuous AI innovation in financial services
Capital One’s approach underscores a broader truth for enterprise technology leaders: driving measurable value with AI requires moving beyond off-the-shelf software toward deeply customized, highly governed architectures. By combining fine-tuned open-weight models, a multi-agent orchestration harness, and proprietary data assets, the bank has established a repeatable blueprint for deploying scalable AI in financial services.
"All of those ingredients were absolutely critical to differentiating in this space and hitting the quality bars as well as the cost and latency thresholds we set for ourselves,” Vanee said.
As the company expands these capabilities across new use cases, its enterprise platform approach helps to ensure that technical breakthroughs translate into safer, faster, and more personalized experiences for its millions of customers.
Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.
Read on the original site
Open the publisher's page for the full experience