Ally Financial Inc. is fully embracing generative AI after the launch of ChatGPT late last year unveiled the technology’s revolutionary potential.
Banks have been leveraging traditional artificial intelligence for years in ways such as predicting market trends and optimizing investment. But generative AI can produce its own data content to increase productivity for financial firms. That has led Ally to create its own generative AI platform — Ally.ai — bringing about 78% of enterprise data across all of the bank’s business lines under the single solution.
Detroit-based Ally (NYSE: ALLY) began experimenting with generative AI last December, said Sathish Muthukrishnan — the bank’s chief information, data and digital officer who is based in Charlotte. In early February, the company had a meeting with Microsoft Corp. (NASDAQ: MSFT) and Amazon.com Inc. (NASDAQ: AMZN) in Seattle. It landed a contract with Microsoft in late April to gain access to its enterprise-grade generative AI software.
Since then, Muthukrishnan said the bank started to build the Ally.ai platform — a proprietary cloud-based tool designed to execute all of its AI-related capabilities. The company launched a pilot of its first use case on June 27. It officially moved to production in late July, with about 700 customer associates using the platform to date.
“I strongly believe generative AI is here to stay, and companies that don’t take advantage of that will be left behind,” he said.
Ally.ai serves as a bridge between internal AI applications and larger external language models such as OpenAI or Google’s chatbot Bard. Muthukrishnan said the platform is fully hosted in Ally’s cloud environment to protect the company’s data. That helps ensure there is no concerns such as AI hallucination, when a language model generates false information, or model drift, the degradation of a tool’s prediction power due to shifts in the environment.
“We believe AI can power the best customer experience in the financial-services sector,” he said.
Ally adopted three principles prior to launching its first use. Those include human intervention, testing the platform with internal customers (employees) first and not disclosing personally identifiable information outside of the bank’s private network. Muthukrishnan said those principles were included in the contract with Microsoft.
The company felt it was best to launch its first use case for its customer care platform. Ally’s customer care associates handle about 10,000 calls a week. At the end of those calls, the agents have to manually summarize the conversation. This is where the Ally.ai platform can hasten that process.
The tool can take audio from a call to generate a transcript along with a summary of the conversation, which saves agents substantial time. Ally associates have the option to correct any mistakes or add additional context. Ally.ai can also generate images from text and convert English to code. Muthukrishnan said text is where the platform “excels today.”
The initial use case proved to be a success. Ally aimed to have 75% of customer care associates describe the platform’s summaries as accurate. It now has 82% giving it a thumbs up. The bank is working on several other use cases, such as integrating the platform in its marketing department.
“Ally wants to be in the forefront, fundamentally,” he said. “The more we use this technology, the better feedback we can provide. … It’s the right opportunity for us to leverage AI.”
Charlotte is a key market for Ally. The bank employs more people here than in its headquarters city, with about 2,800 local employees.
Want to stay ahead of who & what is next? Sent weekly, the Beat is your definitive look at Charlotte’s innovation economy, offering news, analysis & more on the people, companies & ideas driving your Charlotte forward. Follow the Beat
Leave a Reply
You must be logged in to post a comment.