Ask a dev!

With the explosive growth of both digital-first payments and AI, the two are being integrated with exciting results. We are witnessing rapid upgrades to customer experience, for example. As an individual, you can now make payments and get 24/7 support from an AI-based virtual assistant if you encounter a problem.

In this environment, the question of ethical AI is an important consideration when we build and develop these products. ‘Ethical’ AI can mean many things. For Twilio, a vital element when integrating AI with payments is understanding where your data comes from, how it has been handled and how it influences AI decision-making.

This is important for several reasons. First, if the training data contains historical biases and those biases can be replicated or even amplified by AI models, it can be difficult to identify or mitigate these biases. Similarly, in payment systems, errors or opaque decisions can quickly erode user trust.

Secondly, regulators may require detailed documentation of the data and models involved when AI makes payment decisions (e.g., fraud detection, credit scoring, dynamic pricing). Without clear data lineage, it’s difficult to demonstrate compliance or defend decisions.

In short, successful AI integration in payments isn’t just about powerful models—it hinges on disciplined data management. This is why I urge businesses to ask any prospective developer what their communication and payments models are built upon. Likewise, if you’re building paytech tools in-house, striving for unbiased, transparent AI systems is something I strongly advise business leaders to keep in mind.”

There are three key things businesses can do to create great relationships with developers.

The first is to ensure that business leaders clearly communicate the company’s needs. Too often they might say to a developer, ‘I want an AI experience for my customers.’ Great. But if they’re able to come in and say, ‘My business is preparing for a new product launch in Q4, and we need an AI-enabled assistant to launch at the same time to support our customers to make the best use of this new product’—that is a clear ask that allows developers to engage in a much more practical, solution-oriented manner.

The second key thing is to have experts on the business side who understand the company’s own technology. The more context they have about how their own existing technology stack runs and operates—and about the issues they’ve had—the more effective an external developer can be when they come in to help. They can more quickly add value.

The third quality I notice of companies that collaborate successfully with developers is a willingness to experiment. Being able to enter into a dialogue and keep adjusting and working together to find the right solution is an underrated virtue.