May 20

Cloud data platforms with consumption-based pricing like Snowflake are a critical enabler of advanced analytics, but costs can unexpectedly rack up. Capital One’’s Slingshot saves money by monitoring data warehouse performance, making optimization recommendations, and automating governance, and the bank is betting that other firms will be willing to shell out for its battle-tested software. 

Capital One realized that having its different lines of business individually manage their own Snowflake compute requests and configuration led to unexpected costs, so it built and used Slingshot over the last several years to streamline Snowflake’s provisioning processes. It has decreased its cost-per-query by 43% using the tool. 

Slingshot is the first public release of its new B2B software business, but the bank plans to launch other products in the future.  

“We recognize that many other businesses are facing similar data management needs … so bringing some of the tools we’’ve built and scaled to market as enterprise B2B software solutions is a natural evolution for us,”” said the head of Capital One Software, Ravi Raghu, in a press release.  

$4.2 billion ThoughtSpot has spent the last few months successfully using the tool to “optimize usage and control costs,” it said in the release.  Other tools to help businesses manage, monitor, and optimize their Snowflake use include WhereScape, DataDog, and BryteFlow.

The Royal Bank of Canada is using PIN numbers – combined with biometrics and card authentication – on its mobile app because consumers are less likely to give away PINs online versus other personal information that scammers can take advantage of.

As consumer fraud losses climb, RBC is integrating the traditional safeguard of a personal identification number (PIN) into its mobile app authentication process because people are less likely to share it online. 

“A PIN is a well-known asset and not something that consumers are likely to share as much as other information,”” RBC’s chief digital officer Peter Tilton told American Banker. The IT project of implementing PIN on mobile lasted several years, he added. In addition to their PIN, app users will be prompted for either biometric information (facial recognition or fingerprint) or to tap their debit card to their phone.  

Relying on this classic security measure helps protect against organized crime rings, though it is still susceptible to “friendly fraud” where the attacker knows their victim personally and has their PIN. 

Read the full American Banker story here.

Conversational AI and digital customer service can boost customer acquisition and satisfaction while reducing costs, but lack of sophistication or siloed data can make comprehensive deployment a challenge.  

“There are all kinds of communication tools – chatbots and that type of thing – that help you do more at scale, but not do it really well,” Silicon Valley Bank’s Dan Allred said on stage at the Fintech Nexus conference in New York City, referring to digital customer service tech as “a nut that is not going to crack yet.”  

Indeed, chatbots are often fragmented across lines of business or limited in their scope.  

For example, Westpac, a $965 billion-asset bank in Australia, is working with Kasisto to co-develop an orchestration layer  – a “master bot,” if you will – to pull information from all its existing chatbots. There are a slew of other firms working on conversational AI and digital customer service for financial services, including Clinc and Glia, both of which are Insight portfolio companies.

With its hiring spree, Fidelity plans to expand its abilities for storing and trading cryptocurrencies more broadly, firmly staking its ground despite recent price drops.    

In April, Fidelity announced that it would allow people to put bitcoin in their 401(k) accounts later this year, becoming the first major plan-provider to do so. Now, it’s staffing up to build the infrastructure for the custody and trade of Ether, as well as to add 24-hour trading support and tax-reporting tools.  

The subsidiary has around 400 current clients, though customer acquisition has slowed down with the recent cryptocurrency price drop. Still, Fidelity isn’t deterred. 

“We’re trying not to focus on the downturns and focus on some of the long-term indicators,” including client demand, the unit’’s president, Tom Jessop, told the Wall Street Journal. “We are trying to build infrastructure for the future because we measure success over years and decades, not weeks and months.”

Franklin Templeton Investments’ work with machine- learning- startup Fractal demonstrates the benefits of hyper-personalization and contextualization.

Franklin Templeton Investments used a Fractal product dubbed “Advisor Genomics,” which uses pattern-matching algorithms to predict which financial advisors were likely to interact across a given channel. The program helped FTI associates determine which financial advisors to contact, how to contact them, and what to discuss.    

In one year of working with Fractal, FTI identified 13,200 sales leads, which it closed with a 65% success rate, leading to a 26% increase in sales and $600 million new assets generated. 

Fractal also works with Wells Fargo and Google, and raised $360 million earlier this year.