Neural Rails Aims to Revolutionize Credit Union Digital Sales with AI

By John San Filippo

Rick Phillips, founder and CEO of Neural Rails, is spearheading a movement to transform how credit unions engage with members online—moving beyond traditional chatbots to offer a “humanistic” artificial intelligence (AI) sales and communication experience. In a recent conversation with Finopotamus, Phillips emphasized that Las Vegas-based Neural Rails aims to bring the in-branch experience directly to a credit union’s website, providing comprehensive conversational understanding and ongoing dialogue.

The Core Problem with Current Chatbots

Rick Phillips

Phillips highlighted the significant limitations of existing chatbot solutions used by many financial institutions. He explained that current chatbots are “very transactional” and primarily designed to direct users to a page or connect them with a human agent, rather than engaging in a meaningful conversation.

“Chatbots for the most part understand what you’re saying, but it’s very machine learning. So, it’s looking for keywords and then responding on those keywords,” Phillips stated. He added that these systems often try to get users to “click a link and get out of the chat bot” rather than truly assisting them or leading them through a sales process. He likened the difference to a Google search versus a ChatGPT search, noting that current chatbots lack “contextual memory.”

Neural Rails’ Differentiating Approach

Neural Rails distinguishes itself by leveraging advanced AI, Phillips explained. “We absorb your entire universe, and we build a container. So that universe is your website. It could be other information that you would like to provide to us: documentation, special information here, there, social media, whatever you want to provide to us. We build that container, and then we train ChatGPT on that container,” he continued. “This comprehensive ingestion allows the AI to intimately know everything about your credit union, and it can have dialogue accordingly.”

This means the system can answer detailed questions about various topics, from branch hours to complex financial inquiries, all while maintaining the context of the conversation. Phillips noted that if a user asks about auto loans, the system “already knows that that’s the area you’re in. And so, it’s going to drive that conversation with detailed information.” He further emphasized that the AI understands slang and misspellings and won’t get “lost immediately” if a user makes a typo.

Phillips asserted that Neural Rails tailors responses based on the user’s generation, recognizing that “boomers are more likely to be interested in” different things than younger generations, who might be focused on building credit.

Implementation and Integration

The implementation process for Neural Rails is remarkably fast, said Phillips, adding that for the core service, a credit union needs as little as one week to go live. The process involves ingesting the entire website and then running it through a “training harness” where thousands of financial questions are fired at the service to ensure it’s well-versed in the credit union’s specific offerings. The system then ingests information daily to stay up to date with any changes to the website.

Neural Rails also integrates with existing marketing systems like Hubspot. Phillips highlighted that this integration prevents “abandonment” where a user clicks a link from an email and is simply left on a generic webpage. Instead, Neural Rails continues the conversation on the website, leveraging the context of the email to provide a seamless and engaging experience.

Phillips revealed to Finopotamus a significant partnership with Mahalo Banking, a digital banking platform. While the marketing site implementation remains consistent for all credit unions, the integration with Mahalo for digital banking offers even deeper capabilities, enabling “full AI sales and marketing communications site navigation” within the banking platform. This allows the AI to maintain a long-term relationship with members, understanding their spending habits and life events to offer proactive assistance and tailored financial guidance.

Data and Future Outlook

Neural Rails also provides monthly logs of interactions, allowing credit unions to analyze trends, understand what people are asking, and identify where users might be dropping off in processes like loan applications, he noted. This data allows credit unions to “tweak” their offers, processes, or even terminology to optimize engagement and reduce abandonment rates.

Phillips expressed confidence in Neural Rails’ advanced position in the market. He asserted that the company is “legions ahead of the rest of the world” because they have “figured out how to put guardrails around ChatGPT and focus it only at your data.” This means the AI will only provide responses based on the credit union’s provided information, avoiding “world responses” from the broader internet unless specifically allowed.

For its base technology, Neural Rails partnered with Colbert, Wash.-based Omniscia AI. Phillips cited examples of Omniscia AI’s success, including “Ask Jen” for Alaska Airlines, which was the first AI tool to pass a “humanistic” test for AI, and a sales engine for Nordstrom that “outsells salespeople.” According to Phillips, Neural Rails' innovative approach, enabling the “entire loan process to be AI” from initial engagement to application submission, positions Neural Rails to “fundamentally reshape digital sales for credit unions.”