Gridspace software is designed to help businesses analyze sentiment in real time as it was expressed during phone calls with customers. Popular with larger financial institutions, its artificial intelligence-assisted benefits may be useful for financial advisors, too.
Gridspace came from the same lab as Siri, SRI Speech Labs, but unlike virtual personal assistants, Gridspace software was designed to process business conversations between people.
Currently, most Gridspace customers are banks, insurance companies and hedge funds that record conversations. They want to provide real-time insights to their contact center, and their internal communication and trading operation teams, Gridspace CEO Evan Macmillan said.
“We haven’t worked with independent financial advisors or boutique firms yet,” Macmillan told ThinkAdvisor. “That said, I think our software could be useful with those folks in the future. We are still at the dawn of full conversation speech analysis.”
Next year, Gridspace will announce more products for independent financial professionals, Macmillan said.
“Financial advisory relationships are complex and, for now, too complex to be fully automated by software,” Macmillan explained. “That said, Gridspace and other AI software can assist advisors with parts of a financial relationship, namely disambiguating client requests, managing procedural risks and streamlining resolution discovery.”
Some “future-use cases” he identified include:
- Measuring the sentiment of a client’s response to a particular investment idea.
- Chronicling client-advisor interaction histories.
- Screening new financial advisors and products.
“I also think there is value in providing analyses of investment events,” Macmillan said. “This year, we started working with several quant hedge funds that use Gridspace to process real-time financial presentations, earnings calls and news events. I believe financial advisors will soon want these capabilities, too, in order to provide differentiated value to clients.”
Overall, Gridspace helps businesses “automatically document and categorize spoken conversations. The software works by turning speech into highly accurate transcripts and sentiment labels. Most of the speech streams Gridspace processes today, come from phone calls, but the software works with many enterprise voice endpoints,” Macmillan said.
In fact, large financial institutions handle tens of millions of calls a year. It takes hundreds of employees to manually review just a tiny fraction of these calls. Gridspace lets companies scale call review to 100% of calls, so employees can focus on client relationships and high-level operations, Macmillan said.
Such technology could also benefit financial advisors and make them more effective in attracting, keeping and satisfying clients, according to Michael Santoro, a business professor at Santa Clara University.
Yet, looking ahead, there may be some issues about the use of big data AI-assisted techniques.
“There are of course the ongoing debates about whether financial advisors should treat clients as fiduciaries,” Santoro said.
“But regardless of whether the advisor is a fiduciary or has another kind of broker-client relationship, the client still is the ultimate decision maker,” he said.
“In addition to serving customers better, this technology, if not properly deployed, could be abused…. A strong ethical and legal set of controls should be built into the AI system,” Santoro said.
“Big data-driven AI works most effectively when it enhances human action — as opposed to supplanting it.… So, I would say lawyers and ethicists should be involved at the design and implementation stage rather than after the fact to clean up the messes that could arise.”
John Longo, a professor at Rutgers Business School, said that given recent advances in AI, “it is not a surprise that Gridspace is aiming to perform real-time customer analysis.”
“Perhaps the best fit for the financial services industry would be in the compliance space. For example, the software may be able to detect in real time if the advisor was inappropriately guaranteeing high levels of returns or too aggressively promoting certain products,” Longo said. “To be most effective, the software would have to be trained on the voice of the users.
For example, some people are naturally enthusiastic or low key. Unless the software knew this information, it may make incorrect recommendations.” Also, because conversations with clients tend to be sporadic — such as quarterly — the most accurate use of the software is likely with the employees using it on a regular basis, according to Longo.
In addition, financial service firms “should aim to adapt the AI software to anticipate customer concerns or questions and then respond to them in an appropriate, customized manner,” Longo said.
Moreover, Macmillan said one issue that may arise is whether a customer can trust if their advisor is using “fair” tools. “Financial service professionals have a tremendous incentive to maintain the trust of their clients and, accordingly, select AI software that enhances the overall client relationship,” he said.