SoFi Unlocks Coaching Insights and Efficiency Gains with 100% Interaction Coverage via QA Automation

Overcoming the limits of labor-intensive QA

Personal finance company SoFi had humble beginnings by helping individuals reduce student debt and take steps toward financial independence. Today, the company has a customer base exceeding 10 million members, and offers a full range of digital financial products designed to support members at every stage of their financial journey. Over time, these conversations with members became more complex and frequent—making it exceedingly difficult to consistently evaluate interactions and catch potential issues early, especially around calls related to financial hardships or regulatory concerns.

“We were touching maybe 2% of our interactions coming into the contact center,” says Joshua Johnson, Senior Director of Operations Strategy at SoFi. “That’s not enough to identify risks to the business in an effective way.”

To further compound this issue, the QA team also heavily relied on spreadsheets and a fragmented workflow tied to their telephony platform to manually manage their workload and distribute QA evaluations. “It was an extremely time-consuming and tedious process using things like Google Sheets.”

Full visibility into member conversations

To expand its QA coverage and get a clearer picture of what was happening in member conversations across their huge customer base, SoFi brought in Observe.AI to automate its QA process. This move has radically improved the efficiency of assigning and reviewing calls and “helps free up our QA resources to spend our time looking at the risk that’s coming into the business,” according to Johnson.

The QA team now reviews every member call including those that were previously buried in the backlog and generates insights on agent performance based on 100% of all interactions.

“We’re looking to do a full end-to-end workflow buildout for our QA team using automation, so we can evaluate conversations at scale and shift analyst time toward the riskiest interactions,” says Johnson.

Coaching and product insights at scale

SoFi is using insights from Observe.AI to feed findings from call reviews into its coaching program and to guide product decisions based on what its members are actually saying.

As Johnson explains, “We’re taking our agents’ soft skills that are being evaluated on the QA scorecard and sending those insights to our frontline management to be able to coach our agents on better performance. For Product teams, being able to take feedback from what members are saying about our financial products, helping them understand the top five to ten things they can take action on today to improve the product, has made a difference.”

Summarization capabilities also help the QA and Product teams to act quickly. By generating summaries of conversations with members, Observe.AI enables them to identify common pain points, such as confusing product features or frequent complaints, without combing through hours of calls. It “allows them to take action in the right areas,” adds Johnson.

Cutting through the AI noise

With the rapid rise of AI, SoFi focuses on what really drives value for the organization, and that’s AI’s ability to support agents and uncover important business insights. Adopting Observe.AI and being able to unlock efficiency and insights has helped to change this understanding.

“There’s a lot of fear in AI replacing jobs,” says Johnson. “The biggest piece we’re helping our internal teams understand is that AI is an assistant. It’s a tool to help you be better at your job, not to replace your job.”

Building a connected system

SoFi is currently expanding its use of Observe.AI across QA, coaching, and product feedback. By linking these processes, the company is creating a more efficient and connected system that puts insights into action faster.

“We’re looking at coaching workflows coming from those insights,” says Johnson. “It’s all about connecting the dots—QA, coaching, product—in a very automated way.”

His advice for other leaders thinking of adopting AI in their operations is to get the fundamentals right. “Do your due diligence. Make sure you have the right requirements from your business. That’s a key piece.”