December 14, 2018 | Written by: Kathy Tomes
Categorized: AI | Banking | FinTech
Share this post:
At the October Money 20/20 in Las Vegas, Yousef Hashimi, Vice President and Partner of IBM Cognitive Big Data and Analytics, and Rob Palacios, Executive Vice President and Director of Texas Capital Bank (TCB) Labs, sat down for a fireside chat about driving value in organizations with artificial intelligence (AI) and what it takes to be AI ready to reap the benefits of building solutions rather than just buying components of AI technology. Developing trust was a reappearing theme throughout the discussion.
Since the TCB business serves middle market clients and is a commercial-focused organization, it doesn’t rely on brick and mortar stores. Therefore, Palacios says, the bank seeks to use technology in ways to minimize any disadvantage that a lack of physical stores might cause. The first TCB Labs AI pilot focused on finding business customers that fit to the products and services that their bank could provide. Banks often spend a huge lift and amount of time on trying to find the right next client, right next product, and how to serve the client better. So TCB wanted to look at external and internal data performance that would allow them to better source clients and to better identify, rank, and match clients to their portfolio–improving overall customer relationship management.
Palacios was exploring cognitive robotics processing automation (RPA), artificial intelligence, machine learning, blockchain, and fintech partnerships, but he knew the real challenge with the use of emerging technology would be the cultural change that had to happen in the business. Trust was a key component for achieving success on a variety of fronts.
Trust in how AI can provide business value
Palacios explained that stakeholders and employees want to trust that innovation teams are not simply exploring technology for technology’s sake and that the innovation will lead to true business value. If you can’t deliver value, you’re not valuable. Thus, TCB tied their AI project to their five-year business roadmap and the way they make money or reduce a total cost of ownership. Palacios remarked, this alignment “gets more people in the boat with objectives that align with what we’re trying to do and allows for the culture to coalesce.”
Trust in AI expectations
Palacios also noted it was important to set expectations in the AI project early on since people can get really excited about AI and think that it is a silver bullet for all of their problems. He says it is important to help the business understand that while AI will help with the initial lift, it isn’t a silver bullet. He encouraged others not to focus on the AI, but rather the change in user experience that the company is trying to effect. AI is a part of that, but design thinking and data work and other things have to happen to get to the end result the business wants. If teams don’t make it just about the AI, others will lower their expectations that the AI will solve everything. The team needs to explain that it is how the company adopts and rolls out the AI project that will affect the way the company achieves the actual business value.
Trust in the AI data and the innovation approach
Palacios also explained that data itself causes a trust challenge with AI projects. Given the amount and unwieldiness of data, it’s hard to always trust the data. How do you get the right data dependably where you can actually do something with it, with machine learning and cognitive? Palacios highlighted the importance of choosing the right AI partner so TCB could “fail quickly” to get to the right outcome more quickly. He shared that TCB looked at three business models when choosing an AI partner, with the question, “How can I get to the market fastest?” top of mind.
TCB Labs looked at three possible AI innovation business models:
1. Should we hire 20 TCB labs staff: Data scientists, design thinkers, technologists, change management folks? But it would be tough to find the right people to come in at the right time and start driving value immediately.
2. Can we cobble together a lab by picking and choosing people from different organizations and consulting agencies? Yes, but this approach could be expensive and difficult to create synergy among the different parts.
3. What about an already trusted partner? IBM introduced TCB Labs to the IBM Garage concept, a one stop shop for bringing together all the different resources to create an innovation organization on a subscription basis.
TCB chose the last approach and was able to start innovation from day one instead of spending time hiring. Palacios emphasized that design thinking was helpful because they were able to look at business problems first and how to solve them. And since they were able to use a fail fast model, they could move quickly to NOT focus resources where they saw low possibilities of success.
Trust in the AI and human relationship
Hashimi praised TCB for their success and lessons learned and emphasized the importance of considering the psychological impact on humans and their trust in AI. Automation often seeks to eliminate tasks for human beings, and employees fear the loss of their jobs and livelihood. IBM’s approach to AI is not to replace humans, but rather to create augmented intelligence that helps amplify human cognition. Companies should ask: How do you leverage technology to help humans do what they do better? And how can technology bring value in such a way that it allows them to do higher value things?
–>Learn more about IBM Banking and Financial Markets solutions
Watch the full AI fireside chat at Money20/20 below.