Lessons on alternative data and lending with Conductiv

In the early stages of the Covid-19 pandemic, traditional banks encountered a sudden need to digitize. These institutions had to turn to new technologies to replace the in-person services that were no longer available under lockdown. 

To Gopal Swamy, CEO and Founder of Conductiv, this presented an opportunity to not just digitize banks’ lending decision-making processes, but to help them improve  the process and equity in lending by leveraging alternative data sources. 

In a conversation with The Financial Revolutionist, Swamy, along with Zach Lambert, Head of Product at Conductiv, outlines what Conductiv does, and how the team went about building their product. 

What

Conductiv aggregates permissioned-data providers—including account data, income data, utility payment data, static and dynamic identity data—to provide an API and platform for smaller lenders, which they can use to augment lending/decisioning processes. Conductiv also offers comprehensive data-visualization capabilities for lenders 

Why

Credit scores provide an incomplete, and often exclusionary, image of an applicant, and small lenders lack the bigger picture now increasingly accessible to larger players. 

“We're driving efficiency within the bank or the lender, but also driving equity in lending,” Swamy said. “Those are the forces that are coming together.”

Lambert also cited the long-term failure of financial inclusion in the U.S. Conductiv can lessen risk and boost ROI, while bringing more people into the fold. 

How

Conductiv followed a Lean UX approach by presenting a bare-bones MVP to gauge market interest, highlighting the visualization possibilities of their product first and then building out from there. 

The team started just with an API—a single API call to their client, and a single webhook on Conductiv’s side that the client can call back to. “What is the minimal lift for the client so we can integrate quickly and get up in terms of operations?” Lambert asked. 

But Conductiv soon learned that, if you’re talking about an endpoint with a financial institution, then they’re relatively technologically sophisticated. Some banks lack API capabilities, which compelled Conductiv to create a standalone platform that banks can opt into using as well. 

Across the board, Conductiv worked in Angular with Java microservices on the backend, and are now switching from Angular to React as they scale and hire more employees, given that React is a more common language among engineers. 

Conductiv declined to name all the alternative data sources it aggregates, but think of household fintechs like Socure and Plaid. Lambert said they use a champion-challenger methodology (think “king of the hill” but with software) instead of A/B testing to determine the best data source to integrate with.

The team also uses JIRA for product management, Confluence for documentation, Figma for design, and Slack for communication. The team uses AWS for cloud hosting and for pushing their code live. They’re looking into Sagemaker for machine-learning possibilities. 

Looking back

“I wish we had found a way to get more customer feedback more quickly,” Lambert said. “There are ways to get feedback on your product from different stakeholders that aren't necessarily a paying customer yet, but you want to take it with a massive grain of salt, because what you really want to do is value the person who's willing to pay for your product over someone just giving you advice.” Lambert said he’s excited for rapid iteration based on real customer feedback as Conductiv grows. 

“Repeatability: that’s our mantra,” Swamy said. “We focus on repeatable processes; don't do goofy things with the product, no one-off customizations or bad architectural decisions.”