AWS Cost Optimization & FinOps
SageMaker AI Savings Plans deliver up to 64% off SageMaker training, real-time inference, async inference, serverless inference, and processing jobs in exchange for 1-year or 3-year hourly commitment. Compute Savings Plans do NOT cover SageMaker — this is a separate purchase. The break-even is dramatically faster than RI-style commits for steady ML production workloads.
On June 9, 2026 AWS previewed FinOps Agent — a Bedrock-powered agent that investigates cost anomalies via CloudTrail, answers NL cost questions, and opens Jira tickets from Cost Optimization Hub. Free in preview; not a replacement for ownership or tagging.
A lift-and-shift migration copies on-prem specs sized for peak plus headroom, then the migration partner rolls off and nobody owns the bill. The waste is predictable: 30–60% of cost untagged, over-provisioned EC2/RDS, idle NAT Gateways and orphaned EBS, and commitments bought on top of all of it. This is the explicit migration→FinOps handoff — owner first, visibility second, right-size before you commit — with a 30-day checklist and an optimization-backlog CSV.
Tag policies report; only an SCP prevents untagged spend. And ~21% of a typical bill — egress, inter-AZ transfer, Enterprise Support, shared platform services — carries no tag any tool can read. Here is the two-layer enforcement model, the Cost Categories split-charge rules that replace the monthly allocation spreadsheet, and why you show back before you charge back.
Most "RI vs Savings Plan" content treats them as peers. In mid-2026 they are not. A 3-year All-Upfront Compute Savings Plan saves a composite ~$60k/mo SaaS roughly $864k over 3 years vs a 1-year No-Upfront plan—but breaks even at ~22 months and locks instance-family flexibility. Buy the wrong commitment and a Graviton migration strands the discount. Here is the decision tree we use.
Most teams buy Bedrock Provisioned Throughput too early or too late. This is the break-even math — by token volume, by model family, and by traffic shape — that we use in real FinOps engagements to decide which Bedrock pricing mode wins.
An idle NAT Gateway in us-east-1 costs $36.50/month before any traffic: $32.85 in hours plus $3.65 for the public IPv4. A Regional NAT Gateway across three AZs still bills three hours, $109.50 idle.
Cheapest Spot pool is often the unstable one. July 2026: price-capacity-optimized + attribute-based selection, stability scoring, interruption hygiene — before you chase $/hr.
The standard cost optimization checklist no longer cuts it. These 8 modern strategies — from unit economics to automated Savings Plans and cost observability — reflect how engineering teams are actually managing cloud spend in 2026.
AWS Cost Anomaly Detection uses machine learning to flag unusual spending patterns — runaway EC2 instances, unexpected Lambda spikes, or compromised credentials. This guide covers setup, alerting, and automation to prevent bill shock.
Bedrock billing is not a single line item — it is a composition of model invocation costs, Knowledge Base retrieval, Agent orchestration, Guardrails evaluation, and cross-region inference profile routing. Each component has its own pricing model and its own set of cost traps.
AWS Cost Optimization Hub consolidates recommendations from Compute Optimizer, Trusted Advisor, and Cost Explorer into a single prioritized list with estimated annual savings. If you are running three separate cost review processes, this dashboard replaces all of them.
Our articles share what we know. Our consulting engagements apply that knowledge to your specific AWS environment.
We use cookies and similar technologies to analyze site traffic, personalize content, and provide social media features. By clicking “Accept,” you consent to our use of cookies. You can adjust your preferences at any time.
