I work across Go, distributed systems, peer-to-peer networking, and blockchain infrastructure. My focus is simple: build systems that scale cleanly, stay maintainable, and solve real product and platform problems.
If you’d like a more detailed view of my experience, projects, and technical background, you can download the latest version of my resume here.
Backend engineer on Careem's Rides platform (ride-hailing across the MENA region), working across three Go services: the ride-fulfillment core that owns order lifecycle, payments, and provider webhooks; a server-driven UI service that renders the app's live ride-tracking and activity-history screens; and internal API-routing infrastructure. I ship multi-week features as small, independently reviewable stacked PRs, root-cause production issues in payments-adjacent systems, and keep feature-flag lifecycles clean — adding toggles and safely retiring them once rollouts prove stable.
Owned a multi-sprint server-driven UI feature end to end — data ingestion, map markers, a trip-progress "stages" model, ETA-to-next-stop logic, and the post-trip history view — landed as ~8 stacked PRs with the single-stop path kept regression-free.
Built mid-trip cash-to-card banners on the ride-status and payment surfaces, tightly gated on ride state and payment method, reusing an existing feature flag and deeplink helper instead of proliferating new ones.
Root-caused a live ride-creation failure to a character-encoding bug in a deprecated upstream API, fixed webhook idempotency for unknown-ride events, and corrected payment-consent handling for hybrid cash-plus-wallet rides.
Moved service-to-service settlement traffic off the public authenticated path onto private internal routes, gated per-request by a feature flag for a zero-downtime, instantly revertible cutover.
A few projects that reflect my work across backend systems, distributed infrastructure, and practical engineering.
End-to-end server-driven UI feature for rides with intermediate stops — map markers, a trip-progress stages model, ETA-to-next-stop logic, and the post-trip history view — shipped as small stacked PRs.
Event-driven order pipeline in Go over Apache Kafka with Avro and a Confluent Schema Registry — a schema-from-code design where the Go struct is the authority, auto-registered and serialized in Confluent wire format.
Built a production-grade decentralized oracle node in Go, inspired by Ethereum-style client architecture and designed for long-running reliability.
Designed backend systems for a multipurpose co-processor using libp2p and threshold signatures, with a strong focus on coordination and scale.
Explored decentralized storage and lookup patterns through a libp2p-based distributed hash table as part of practical systems experimentation.
I occasionally write about systems, Linux, web development, and ideas that caught my attention.
Loading latest posts...