Ashans
Khadka

I led a one-month build of an in-house deletion API, removing six-figure annual licensing costs.

Regulatory deletion reaches into every store that holds customer data. A request is not done until we can verify its result across those systems. The existing vendor workflow was expensive and put a core legal obligation outside our control, so I built the in-house replacement.

The requirements kept changing as new stores and retention rules arrived. I separated the business rules from the storage layer, which meant we could adapt the underlying systems without rewriting the core logic.

What I'd do differently:I would define, validate, and version the data contract on day one. The first version relied too much on one system's document shape, which made later changes harder to reason about.

MongoDB upgrade leadership across seven environments.

I led the 4.4 through 6.0 upgrades and provided technical leadership for the 8.0 work. Database upgrades have enough ways to fail without discovering one during the change, so I used maintenance windows, standby-cluster safety nets, and rollback paths we had tested before relying on them.

I also wrote reusable runbooks and brought teammates into the work so the knowledge could spread instead of becoming a bottleneck.

Validated change-data reliability.

A dashboard can look healthy while a data pipeline is still missing records. I built validation that compared change events with landing data so we could see whether the path was actually complete.

It showed 100% agreement, and I presented the result to the wider engineering organization.