Personal details
Adnan A.
Technical skills
About
Senior Backend & Platform Engineer with 5+ years of experience delivering production systems with Node.js, TypeScript, Nest.js, Python/FastAPI, AWS, Kubernetes, Postgres, Redis, and distributed workflows. Strong in backend architecture, service decomposition, performance optimization, observability, CI/CD, and cloud cost optimization. Built and operated systems handling 500–800 RPS, reduced MTTR from ~45 minutes to under 10, cut query latency from 30s+ to under 1s, and reduced AWS spend by ~25%. Available for remote backend, platform, and cloud engineering work.
Work Experience
Internal B2B/G2G platforms for international NGOs; Scale: 50k–100k DAU, 500–800 RPS peak, 99.9% availability target.
• Decomposed a monolith into 6 Docker/EKS microservices, cutting deploy time from hours to minutes with zero-downtime releases.
• Implemented Temporal workflows with idempotency/compensation to cut failed cross-service ops ~80% (150/day to <30).
• Reduced core query latency from 30s+ to under 1s through indexing, connection pooling, and service decomposition (Postgres/Redis).
• Reduced payment failures from ~5% to 2% by rebuilding retries and rollbacks in AWS Step Functions.
• Built GitHub Actions + ArgoCD CI/CD with security gates and canary/blue-green releases, supporting 99.9% availability.
• Instrumented Jaeger tracing and Prometheus/Grafana dashboards to cut MTTR from ~45 minutes to under 10 by isolating cross-service latency.
• Stabilized 500–800 RPS burst windows with rate limiting, queue-based smoothing, and graceful degradation.
• Reduced AWS spend ~25% via EKS node rightsizing, autoscaling threshold tuning, and spot for non-prod without SLO impact.
• Delivered backend services for School Management System, Rawalpindi Women University, and Capital Development Authority portal, building APIs and database workflows.
• Cut search latency by eliminating N+1 queries and introducing aggregated queries with Elasticsearch + Redis caching.
• Maintained notification delivery during downstream failures by implementing a Kafka-based pipeline with retries and circuit breakers.
• Prevented traffic to degraded pods by operating Kubernetes/Helm deployments with readiness probes and health gating.
• Secured healthcare data pipelines by implementing AWS Glue ETL with encryption at rest/in transit and fine-grained IAM controls.