Artificial Intelligence at Edward Jones – Two Use Cases

​Edward Jones is a U.S.-based financial services firm focused on individual investors and small businesses. The company reported $16 billion in revenue for 2024, with fees contributing nearly $13.2 billion, and net income reached to $1.98 billion. The firm ranked No. 260 on the Fortune 500, based on its $16.3 billion in revenue. It has more than 20,000 financial advisors throughout North America serving more than 9 million clients with a total of $2.2 trillion in client assets under care as of March 28, 2025.

Edward Jones Ventures, the firm’s strategic investment arm, launched in 2025, and TIFIN Studios provided funding to Grantd, an AI-powered equity compensation planning platform, in September 2025. This investment supports Grantd’s enhancements in automated modeling, alerts, and client reporting for financial advisors managing stock options and RSUs.

This article explores two business use cases of AI at Edward Jones:

• Scaling modernization efforts with generative AI (GenAI): Automating legacy system modernization and streamlining engineering workflows with GenAI to boost speed, quality, and cost efficiency.
• Delivering real-time advisor intelligence: Delivering real-time, contextual insights directly within advisor workflows to enable faster and actionable outcomes across complex systems.

Scaling Modernization Efforts with GenAI

Maintaining legacy systems remains a significant challenge for large organizations.  A 2019 U.S. Government Accountability Office (GAO) report found that federal agencies spend roughly 80% of their IT budgets on operating and maintaining existing systems, many of which are decades old, run on unsupported hardware, and carry known security vulnerabilities.

Among the 10 most critical legacy systems the GAO analyzed, system ages ranged from 8 to 51 years, with multiple agencies lacking complete modernization plans, increasing the risk of cost overruns, delays, and operational disruption.

These challenges are not unique to government agencies. Many enterprises face similar struggles with aging technology that slows modernization, consumes engineering capacity, and introduces operational risk. Edward Jones provides a compelling example of how a company confronted these issues head-on, leveraging GenAI to modernize critical systems, improve productivity, and reduce the risks inherent in legacy infrastructure.

In a conversation with The Wall Street Journal, Kevin Adams, Edward Jones’s chief information officer and head of digital, described the complex environment in which the company’s engineering teams operate. He attributed this complexity to the firm’s long-standing reliance on legacy mainframe systems and its ongoing cloud migration efforts. Adams went on to outline several key challenges:

• Maintaining and evolving legacy mainframe systems that had long served as the backbone of data processing, amid declining specialized expertise.
• Modernizing and integrating legacy platforms with newer technologies.
• Migrating core applications to the cloud while managing architectural complexity.
• Adams strongly implies that addressing data latency concerns, particularly for applications requiring real-time or near-real-time access, necessitates careful workload placement and hybrid approaches.

Hence, Edward Jones partnered with EPAM Systems to integrate GenAI into its software engineering workflows. EPAM provided data and AI capabilities tailored to Edward Jones’ tech stack and workflows, focusing on scalable platforms and repeatable patterns to reduce development risk and speed market delivery.

In the same interview, Kevin shared that the company uses GenAI to support its engineering teams by automating several time-consuming software development tasks, including code migration, refactoring, unit test generation, and documentation. These capabilities were delivered through AI tools and data platforms provided by EPAM.

GenAI was also applied to application modernization efforts, particularly in analyzing legacy codebases. In one instance, the company used GenAI to examine a legacy fee-billing system for retirement invoicing, extract underlying business rules, and generate detailed technical documentation. AI-generated unit tests were used to validate that newly developed systems behaved consistently with the original applications.

In addition, the company described using GenAI to support developer learning and innovation. Engineers relied on AI tools, including an internally developed AI copilot, to scaffold code, optimize implementations, and address complex technical challenges, even when working with unfamiliar programming languages and frameworks.

Per the WSJ article, the company reported that the use of GenAI translated directly into measurable productivity gains, faster modernization cycles, and meaningful cost savings. Kevin shared the following outcomes for the company:

• 20% increase in engineering efficiency, driven by automation of code migration, refactoring, testing, and documentation.
• 98% code completeness and 99% fidelity in modernized systems compared to legacy applications.
• Over 800% acceleration in code delivery within three months.
• 84% reduction in manual review time, lowering engineering overhead.
• $650,000 in cost savings, achieved through faster delivery and reduced manual effort.

Delivering Real-Time Advisor Intelligence

Data silos impose severe limits on decision-making by fragmenting client, market, and operational insights, preventing real-time integration essential for advisors during conversations. Across sectors, the volume of data organizations collect continues to grow rapidly, but the ability to convert that data into timely, actionable insights has not kept pace.

A recent O’Reilly Media report in partnership with dbt Labs highlights the scale of these challenges. The report describes analytics teams in large organizations working across an average of 400 data sources, with nearly 1 in 5 juggling more than 1,000 distinct sources. Over 70% of data teams rely on 5–7 different tools to manage core workflows, while about 10% use more than 10 tools, creating significant integration and operational complexity.

83% of organizations report suffering from data silos, and 97% say these silos negatively affect performance, slowing insight delivery and limiting timely access to critical information. This fragmentation stems from persistent data silos and complex integration challenges that trap information in isolated systems, preventing it from flowing where it’s needed most.

Edward Jones identified a growing gap between the volume of data it possessed and the real-time insights its advisors could access during client conversations. While the firm worked with petabytes of data across hundreds of systems, much of this information remained siloed, limiting its usefulness when advisors needed it most.

These challenges were publicly discussed at the 2024 Databricks AI Summit in a panel discussion. Lindsey Turner, Principal in Data Science, Analytics, and BI at Edward Jones, took to her LinkedIn to share some insights from the panel. Details from that discussion were later reported by RTInsights, which provides further insight into how Edward Jones approached the use case.

The same article mentions that Edward Jones faced several interconnected challenges that constrained advisor effectiveness:

• Client, market, and operational data lived across Hadoop environments, Oracle data warehouses, and legacy systems, making integration and real-time access difficult.
• Advisors needed timely, relevant insights during client conversations, not after.
• The firm’s internal knowledge base, including more than 67,000 pages of content and decades of advisor experience, was challenging to search and apply in real time.
• Advisors needed better guidance on what information to collect and how to refresh client data continuously.

To address these challenges, the RT Insights article mentions that Edward Jones invested in a real-time AI strategy focused on what it calls “advisor intelligence”. The company built AI-powered systems to surface the right insights at the right time, in the appropriate context, directly within advisor workflows. Key elements of this approach included:

• AI-powered advisor copilots that support advisors during live client conversations.
• Enterprise search enhanced with AI-generated answers, improving access to internal content such as JonesLink’s 67,000+ pages.
• Real-time nudges and next-best-action recommendations to guide advisor-client discussions.
• Continuous data integration across hundreds of systems, enabling near-real-time insight delivery.

In a LinkedIn post commenting on the RTInsights coverage, Peter Hegyi, Director of Financial Services at Mulesoft, highlighted how Edward Jones’ advisor intelligence strategy was enabled by MuleSoft’s ability to connect data across more than 900 applications.

The post emphasized that as Edward Jones advances its agentic AI transformation, a critical requirement is not just generating answers through AI, but triggering the right actions across downstream systems.

While specific results for Edward Jones are not publicly available, research from MIT Sloan Management Review highlights that wealth management firms using AI to support advisors have seen measurable improvements in efficiency, client engagement, and decision-making. Real-time AI can help organizations streamline workflows, provide timely insights, and drive actionable outcomes across complex enterprise systems.