- You’ve led tech refresh and database migration work touching close to 200 engineers globally. When a modernization program is that large, what tends to derail it that leadership at the top of the org doesn’t see coming?
There can be scenarios where system integrations are not migrated or implemented correctly, putting both the legacy and modern environments at risk. These integration gaps can create unexpected failures, data inconsistencies, operational issues, and instability across both platforms.
The dependency landscape is often far more complex than leadership initially anticipates. Hundreds of applications may depend on a single database, or a single application may rely on numerous upstream and downstream systems. These interconnected dependencies can make even seemingly straightforward modernization efforts highly complex.
Attention to every detail is critical to the success of these projects. Understanding dependencies, validating integrations, identifying hidden coupling, and thoroughly testing end-to-end behavior are essential to ensuring a smooth transition and protecting the stability of both legacy and modern systems.
- Payments and card processing systems carry uptime and compliance constraints that most software doesn’t. How does that change the way you evaluate emerging tools, especially AI and ML offerings, before they get anywhere near production traffic?
AI can significantly accelerate the development and modernization of systems—helping teams build, upgrade, test, and iterate applications faster.
However, when it comes to production, human cognitive judgment remains critical. Productionizing a system requires context, experience, risk assessment, and the ability to make nuanced decisions about reliability, security, business impact, and operational readiness.
- There’s a lot of noise right now about AI reshaping fintech. From where you sit, where is it actually showing up in payments engineering day to day, and where do you think the hype is running ahead of what the technology can deliver?
From where I sit, the biggest distinction is between AI that improves the engineering around payments and AI that changes the payment decision itself. The first is already delivering value. The second is where I think the industry sometimes gets ahead of itself.
Where it’s actually showing up :
- Engineering productivity is probably the most immediate win. The value isn’t necessarily “AI writes the application”; it’s that a 20-year-old Fintech codebase becomes easier for an engineer to navigate and change.
- Operations and engineering support are quietly valuable.
- Payment Optimizations Risk controls, Fraud and risk decisioning and many other optimization areas.
Where I think the hype gets ahead:
AI directly operates the entire Fintech world.
There is a huge difference between AI autonomously deciding what action to take vs AI recommending which route might produce the best authorization outcome. AI is probabilistic and can produce different decisions under similar circumstances. This should not be handed over to LLM.
- You’ve planned multiyear roadmaps across eight teams. How do you build a roadmap that survives contact with shifting business priorities, regulatory changes, and the pace of new tooling in the AI and cloud space?
I allocated 70% of our capacity to business priorities and 30% to regulatory changes, technical uplift, and other critical project-related enhancements. While there may be minor deviations from time to time, these can be managed and adjusted within the overall allocation.
This approach provides a healthy balance between delivering on business priorities and staying ahead of regulatory and technical requirements, while also reducing last-minute surprises, unplanned demands, and reactive work.
- You’ve held technical lead, people manager, and backlog owner responsibilities on the same projects. What does that combination teach you about engineering leadership that pure management or pure IC paths miss?
Relying purely on management can sometimes lead to an incomplete understanding of how an application operates in practice. This can result in inaccurate capacity planning and, ultimately, unnecessary pressure on engineering teams to deliver within aggressive timelines.
On the other hand, relying purely on an Individual Contributor (IC) perspective can miss broader considerations such as strategic planning, cross-team dependencies, prioritization, stakeholder expectations, and other key areas beyond day-to-day application operations.
I believe having both perspectives (strong management oversight and deep IC-level technical expertise) creates a much more balanced approach. This combination gives me greater confidence in capacity planning, decision-making, risk management, and execution, while making it easier to navigate challenges and successfully deliver projects.
- For engineers earlier in their careers who want to grow into roles like yours inside large financial institutions, what skills do you think are undervalued right now, and what do you wish more people coming up understood about working inside regulated environments?
Cognitive abilities, emotional balance, and strong interpersonal skills are often undervalued in highly regulated environments, yet they are essential to performing effectively in these settings.
People need to understand the importance of timelines and commitments while also maintaining the right balance to deliver quality outcomes without creating unnecessary pressure. Equally important is developing a habit of continuous learning—setting aside time to understand the applications, processes, and business areas they support.
Consistent learning builds deeper expertise, improves decision-making, and helps engineers anticipate risks rather than simply react to them. Every new insight, experience, and lesson learned contributes to greater system stability and strengthens the individual’s ability to excel in their role.
