Database Cost & Reliability Review

The problem

Your database bill keeps climbing, but nobody can say precisely why. Instances were sized for a launch two years ago. Storage and I/O charges grow quietly. Replicas were added "to be safe." Somewhere in there is real waste — and usually, right next to it, a reliability risk nobody has measured: a replica that can't actually fail over, a backup that's never been restore-tested, a query pattern one traffic spike from an incident.

What we do

A senior database engineer reviews your environment end to end — sizing, storage, I/O, connections, replication, backup/recovery, and the query patterns driving load. We work read-only against replicas, snapshots, or read-only credentials. The output is a ranked list of findings: where the spend goes, what it would take to recover it, and what risks to address before they become incidents.

What you get

• A written report with an executive summary a non-DBA can act on.
• A prioritized findings list, each with evidence and an impact/effort rating.
• A cost savings section with the math behind each opportunity, sized to your actual usage — not a generic percentage.
• A reliability risk section covering replication, failover, backup/recovery, and load risk.
• A 30/60/90-day action plan sequenced by impact, effort, and risk.
• The diagnostic queries we ran, so your team can re-run them later.

How it works

1. Scoping call (30 min) — confirm engine, scale, access model, and tier.
2. Read-only access — replica, snapshot, or read-only role + metrics export, under a mutual NDA.
3. Diagnostics & review — we run the analysis; a senior engineer interprets it.
4. Report + walkthrough — written report plus a live readout call.

Engagement tiers

Tier	Scope	Deliverable
Lite	One database/cluster, top 5–7 findings	Concise written report + query pack
Standard	~3–5 databases or one platform, top 10 findings	Full report, 30/60/90 plan, live readout
Deep	Full estate + remediation design	Full report + remediation plan + advisory sessions

Engines & platforms

PostgreSQL, Aurora (PostgreSQL/MySQL), MySQL, RDS. AWS primarily; GCP/Azure equivalents on request.

What this is not

Not a managed service, not 24/7 on-call, and not a write-access remediation engagement — we hand you the plan and advise; your team executes. We don't promise a specific savings figure before reviewing your data.

How AKS works with your data: read-only by default — replicas, snapshots, or read-only credentials. Mutual NDA, documented access and data-handling process, access removed on delivery. You own every artifact we produce.