Real-time systems that don’t lose data.
Senior software engineer, 8+ years. I build telemetry pipelines and event-driven backends that stay correct under load — and ship them to enterprise clients.
ABSTRACT · Eight-plus years across telemetry, logistics, and enterprise platforms. At Tenderd (YC S18) I led a 7-engineer team to deliver the module that onboarded Saudi Aramco inside a hard <2.5-monthwindow, and migrated live CDC → Kafka telemetry across 200+ devices with zero event loss. Earlier: event-driven backends serving 5M+ users at Bykea.
Field notes, written in production
Migrating IoT Telemetry Ingestion to a New Schema with Zero Event Loss
A third-party device server that streams telemetry from the fleet was moving to a new cloud version. The change repointed writes from the existing location/alarm tables to new v2 tables. We ingest that data via CDC → Kafka, so the cutover risked dropped or duplicated pings across every device on the platform.
Δ = 0 across 200+ live devices →ENTRY 02 · 2024DELIVERY · ENTERPRISEDelivering an Equipment-Booking Module to Onboard Saudi Aramco
Saudi Aramco needed an internal equipment-reservation system: departments reserve equipment, use it, and return it by a deadline so the next department can book it. It had to be delivered on a hard timeline as the condition for onboarding them onto the platform.
delivered in <2.5 months · ~200 assets managed →Selected builds
Fleet Telemetry & Operations PlatformReal-time telemetry, timesheets, and safety metrics for heavy equipment fleets.
2025 · TenderdP·02Music Distribution & Analytics PlatformA Go backend for cross-platform music release and track resolution.
2024 · ClozerP·03High-Scale Logistics BackendEvent-driven services and support tooling for a 5M+ user platform.
2022 · BykeaNotes on how I design
Event-Driven Backbones with KafkaWhen to reach for Kafka, what the outbox pattern really buys you, partitioning for ordering, and the things that break in year two.
11 min · read§ 2Designing a RAG Pipeline for ProductionChunking, hybrid retrieval, reranking, grounding with citations, and the evals that separate demo-ware from production.
13 min · read§ 3Multi-Tenant Isolation PatternsPooled vs siloed vs cells: picking an isolation model that matches your blast-radius budget, not the hype cycle.
9 min · read§ 4Observability That People Actually UseSLOs, burn-rate alerts, and why your dashboard graveyard is a product problem, not a tooling one.
8 min · readReliability work, taken on commission.
Kafka audits, telemetry ingestion builds, enterprise module sprints, and fractional architecture. Fixed scope, written findings, production-grade delivery.