Meet Manish, AI Lead at Crosskey

Tell us a little about your background

I studied mathematics as an undergrad and followed that with a master's in computer applications. I joined Crosskey in 2016 as a consultant and what started as a short assignment turned into a decade-long journey through several roles: Senior Developer, Platform Manager, System Architect, and now AI Lead. Before Crosskey, I spent two years at Nordea Bank in the Asset and Wealth Management area.

I did not arrive at AI from the outside. I built and architected real-time payment systems, designed cloud architectures, embedded PCI-DSS and OWASP frameworks into production, and sat in Security, Architecture and API governance forums long before AI became a strategic priority. That background shapes how I approach AI. I know how banking systems work and what can go wrong.

There's pressure across the industry to move fast with AI, what does leading AI adoption in banking actually mean in practice?

Speed is a capability, not a goal. The pressure is there, but in banking it is worth asking what you are actually trying to achieve and whether speed is even the right question.

What concerns me more is that speed scales whatever you already have. Weak processes, unclear ownership, informal oversight automating those do not fix them, it just moves the same problems faster. And when AI takes over parts of a process, human judgement that was built in does not automatically get replaced. It often just disappears.

Leading AI adoption in banking means being deliberate about that before anything gets automated. Understanding which parts of a process rely on informal oversight, experience, intuition, someone slowing things down when something feels off, and deciding whether that can be formalized first. That is the actual work.

At Crosskey that is why we run AI forums where we ask these questions before a process is touched. It is also why our AI tooling rollout was phased and structured rather than open access from day one. Trust has to be built into the process before scale. Once it is, you can move fast. But the order really matters.

How does your role at Crosskey contribute to client success?

The most direct way is making sure the AI we build and adopt does not create any risks for our clients. Every initiative I lead has to be production-safe, audit-ready, and aligned with the regulatory environment our clients operate in.

I am also focused on embedding AI into how we build software and eventually into the banking products themselves. When we move faster with higher quality and lower overhead, that value reaches clients too.

How do you stay motivated in your work?

The AI space is moving so fast, and that keeps me motivated. What I find particularly interesting is the gap between what AI can do in theory and what it can actually do in a regulated banking environment. Getting AI to work in a way that satisfies developers, compliance, security, and leadership at the same time is a genuinely difficult problem.

The other part is watching colleagues who were skeptical about AI become confident users of these tools. Adoption only happens when people trust both the technology and the process around it. That takes time, but it is one of the most satisfying parts of the job.

‍