Ankit Rawat On Building Trustworthy AI Systems

Ankit Rawat, a Senior Software Engineer at Meta Platforms, has spent 13 years building payment and logistics systems at top companies like Amazon, Wayfair, and Meta. He emphasizes the importance of production-ready AI requiring observability built in from day one.

On March 25, 2026, the UK's Financial Conduct Authority published its Payments Regulatory Priorities report, highlighting the need for governance in agentic AI payments. This admission underscores the gap between capability and trustworthy production deployment in the payments industry. According to Rawat, 99% of companies plan to put AI agents into production, but only 11% have done so, citing implementation challenges around data, governance, and security.

Rawat's career has been built on the conviction that if you cannot see a system break, you cannot fix it before someone's money disappears. At Meta, he developed an observability methodology that now spans 100+ teams and underpins payment flows serving hundreds of millions of users. This approach focuses on instrumenting real user journeys instead of just infrastructure.

Most teams, Rawat notes, are watching the wrong thing. Their dashboards show healthy infrastructure, uptime, latency, error rates, and everything looks fine right until a user is already affected. However, dashboards can't show what users experience, such as the payment step that keeps stalling or the flow where people quietly drop off.

Rawat's work at Meta started as a four-team initiative and expanded to 100+ teams, directly surfacing a multi-year payments opportunity that existing tooling had never detected. This experience highlights the importance of system visibility and the need for engineering teams to think differently about observability.

In the payments industry, the gap between capability and trustworthy production deployment is not theoretical, but a real issue where failures happen. The technology is ahead of the governance, and the governance is still ahead of the engineering discipline required to make it trustworthy in production.

Rawat's insights emphasize the need for a new approach to building trustworthy AI systems. By prioritizing observability and instrumenting real user journeys, companies can close the gap between capability and trustworthy production deployment.

The importance of trustworthy AI systems cannot be overstated, especially in the payments industry where financial events can have significant consequences. As regulators scramble to govern technology that is already ahead of them, companies must prioritize engineering discipline and observability to ensure that their AI systems are production-ready and trustworthy.

In conclusion, Ankit Rawat's experience and insights highlight the need for a new approach to building trustworthy AI systems. By prioritizing observability and instrumenting real user journeys, companies can close the gap between capability and trustworthy production deployment, and ensure that their AI systems are production-ready and trustworthy.

The implications of this are significant, not just for the payments industry, but for any company looking to deploy AI systems in production. As the use of AI becomes more widespread, the need for trustworthy and observable systems will only continue to grow. Companies that prioritize engineering discipline and observability will be better equipped to deploy AI systems that are reliable, secure, and trustworthy.

Ultimately, the key to building trustworthy AI systems is to prioritize observability and instrumenting real user journeys. By doing so, companies can ensure that their AI systems are production-ready, trustworthy, and capable of delivering reliable and secure results.

In the end, it is clear that building trustworthy AI systems requires a new approach, one that prioritizes observability and engineering discipline. As the payments industry continues to evolve, it is essential that companies prioritize these principles to ensure that their AI systems are reliable, secure, and trustworthy.

The future of AI depends on it.