You need an AWS-hosted relational database that ingests CSVs on its own and serves clean, fast SQL views for reporting — not a pile of one-off scripts someone has to babysit. That’s the kind of work I’ve shipped in production.
At the Department of Science and Technology, I ran RDS as the core database behind containerized services on AWS, optimized slow queries through root-cause analysis, and automated the build/deploy path with S3, CodeBuild, and CloudFormation (Infrastructure as Code — the “plus” you listed). On Aqualytix, a live PostgreSQL monitoring platform, I wired S3 for storage and CloudWatch for metrics, logs, and alarms with SNS email alerting — exactly the backup/monitoring posture your scope calls for.
Here’s how I’d approach the CSV-to-reporting flow:
I’d start by deploying RDS and the schema (indexes, views), then layer in the automated S3 ingestion, then backups/IAM access controls and CloudWatch monitoring — each piece deployable and demoable before the next.
Want to hop on a quick call so I can hear your data shape and CSV format, and sketch the schema with you?