Whenever you're ready, walk me through how you'd approach this.
How would you approach optimizing a data processing workflow?
Describe a practical method for profiling the workflow, locating bottlenecks, and selecting improvements across ingestion, transformation, storage, and orchestration. Explain how you would preserve correctness while improving runtime, throughput, cost, and data freshness. Include how you would validate changes, handle failures, and monitor the optimized workflow in production.