Mark Wlodawski

About

Backend/data engineer focused on AWS-native Python systems: built a FastAPI microservice on ECS/Fargate serving real-time analytics at millions of daily requests with strong reliability (OAuth2/JWT, retries/timeouts, correlation IDs) and autoscaling. Also delivered Glue/PySpark ETL pipelines to curated S3 Parquet/Athena with schema evolution + data quality controls, owned Airflow pipeline incidents, and has a track record of measurable performance and cost optimizations (e.g., ~80%+ query latency reduction; reduced logging/NAT/Fargate spend).

Experience

Education

Key Strengths

• Designed and operated stateless FastAPI microservice on AWS ECS handling millions of daily real-time analytics requests
• Production-grade reliability patterns: OAuth2/JWT auth, strict timeouts, exponential-backoff retries, centralized sanitized error handling with correlation IDs
• Scalability engineering: ECS autoscaling (CPU/latency), Redis caching, DB query optimization/connection pooling, offloading work to Kafka/Celery
• Infrastructure automation: Terraform IaC + GitHub Actions CI/CD (Docker to ECR, rolling deploys)
• Cloud cost optimization: reduced CloudWatch logging costs, right-sized Fargate tasks, minimized NAT egress via VPC endpoints; implemented budgets/alarms and cost dashboards
• Built AWS Glue ETL to curated partitioned Parquet on S3 with Athena consumption; managed schema evolution and data quality gates
• Incident ownership for batch/data pipelines: detected via SLA/metrics, triaged via Airflow logs, mitigated with rollback/backfill, prevented recurrence with schema validation and quarantine path
• SQL tuning with measured impact: reduced Postgres endpoint latency from ~2–3s to ~300–400ms via CTE rewrite, composite index, and keyset pagination