Amazon SRE Capacity Planning Interview: A Real Case Study from AWS

The Amazon SRE capacity planning interview is a gatekeeper, not a test of raw knowledge. It separates candidates who can translate metrics into business‑critical decisions from those who simply recite textbook definitions.

TL;DR

The interview consists of a five‑round process, culminating in a 30‑minute capacity exercise that probes signal‑to‑noise judgment. Candidates who surface with a clear capacity‑signal framework win, while those who focus on memorizing AWS limits lose. Salary after an offer typically lands between $170,000 and $210,000 base, with 0.05 %–0.1 % equity and a signing bonus that ranges from $20,000 to $45,000.

Who This Is For

This article is for senior‑level SREs who have at least three years of production‑scale experience on AWS, are currently earning $150K‑$180K base, and are targeting a move to Amazon’s SRE organization. You likely have shipped services with >99.99 % uptime, own capacity dashboards, and are frustrated by interview loops that reward buzzwords over concrete trade‑offs.

What does the Amazon SRE capacity planning interview actually evaluate?

The interview evaluates a candidate’s ability to synthesize telemetry, forecast demand, and articulate risk in business terms. In a Q2 debrief, the hiring manager pushed back on a candidate who listed “CPU‑to‑memory ratios” because the signal was irrelevant to the product’s SLA. The judgment signal is the candidate’s capacity‑signal hierarchy, not the depth of technical jargon. The first counter‑intuitive truth is that the problem isn’t your answer — it’s the way you prioritize the data you present. Not a list of metrics, but a narrative that ties traffic spikes to revenue impact, wins the round.

The interview also tests organizational psychology: can you convince a senior TPM that a 20 % headroom is a safety margin, not a wasteful over‑provision? The second insight is that seniority is measured by the confidence to challenge assumptions, not by the volume of equations you can write on a whiteboard. Candidates who treat the exercise as a “trick question” lose, while those who treat it as a “business conversation” succeed.

How is the interview structured and what timeline should candidates expect?

Amazon runs a five‑round interview loop lasting an average of 14 days from the first screen to the final debrief. The first two rounds are 45‑minute phone screens focusing on systems design and incident response. The third round is a 30‑minute capacity planning exercise delivered via shared screen. The fourth round is a deep dive with a senior SRE who probes your trade‑off reasoning. The final round is a culture‑fit discussion with the hiring manager and a TPM.

The timeline is intentionally tight to surface “real‑time decision making.” Not a marathon of endless follow‑ups, but a sprint where you must produce a capacity model under pressure. In a recent interview, the candidate received feedback on the capacity exercise 48 hours after the call, giving them a single day to iterate on the model before the next round. The judgment you must make is whether you can iterate quickly without compromising rigor.

Which signals separate a competent SRE from a superficial candidate?

The core signal hierarchy consists of three layers: (1) demand quantification, (2) supply articulation, and (3) risk communication. In a debrief after a June interview, the hiring manager highlighted that the candidate who nailed the exercise presented a “demand curve” derived from a 90‑day traffic sample, then overlaid a “service‑level budget” that directly referenced the product’s revenue target. The candidate’s risk communication was framed as “if we under‑provision by 10 %, we risk $1.2 M in lost transactions per quarter.”

The third counter‑intuitive insight is that the problem isn’t the model’s precision — it’s the story the model tells. Not a perfect Monte Carlo simulation, but a clear, actionable recommendation. Candidates who bring a spreadsheet full of confidence intervals but cannot link those numbers to business outcomes are flagged as “data‑rich, insight‑poor.” The judgment is to prioritize business impact over statistical elegance.

What concrete scripts can a candidate use to demonstrate depth in the capacity exercise?

Script 1 – Opening the exercise: “Based on the last 90 days, we saw a 30 % weekly growth trend; projecting that forward, I’d expect a 2.4× increase in traffic over the next quarter.” Script 2 – Responding to a pushback: “I understand the concern about over‑provision; however, the cost of a 5 % headroom is $12 K per month, while the cost of a 10 % SLA breach is $1.4 M in lost revenue, so the ROI on the headroom is 117×.” Script 3 – Closing with risk framing: “If we allocate 15 % more capacity now, we can absorb a three‑sigma spike without triggering the auto‑scale throttle, keeping the SLA at 99.99 %.”

These scripts were used verbatim by a candidate who later received a $190,000 base offer. The judgment you must make is to own the narrative, not to wait for the interviewers to ask the right question. Not a defensive posture, but a proactive stance, changes the interview’s direction.

How should a candidate negotiate compensation after surviving the interview?

Negotiation begins once the offer includes the base, equity, and sign‑on numbers. In a Q3 debrief, the hiring manager disclosed that the candidate’s equity grant was initially set at 0.04 % but was increased to 0.07 % after the candidate referenced a comparable senior SRE at a rival cloud provider earning 0.09 % equity. The judgment is that you must anchor the conversation on market data, not on vague “fairness” arguments.

The fourth insight is that the signing bonus is not a goodwill gesture; it’s a lever to close the gap between your current compensation and the offer. Not a request for a higher base, but a request for a higher sign‑on that you can accept without tax penalty. A typical successful script is: “Given my current total compensation of $250K, I would need a $30K signing bonus to bridge the difference and feel comfortable transitioning.” The hiring manager respected the data‑driven approach, resulting in a $175,000 base, $38,000 signing bonus, and 0.08 % equity package.

Preparation Checklist

Review Amazon’s three‑layer capacity‑signal hierarchy and rehearse articulating each layer in under two minutes.
Build a capacity model using a real AWS service (e.g., DynamoDB write capacity) and practice explaining the business impact of each metric.
Conduct a mock interview with a peer who plays the role of a senior TPM, focusing on risk communication.
Study the compensation benchmarks for senior SREs on Levels.fyi and prepare a concise equity comparison table.
Work through a structured preparation system (the PM Interview Playbook covers capacity‑signal framing with real debrief examples).
Prepare the three scripts listed above and memorize the exact phrasing.
Set a timeline: complete the mock loop within seven days, then schedule a final review two days before the interview.

Mistakes to Avoid

BAD: Listing every AWS limit you know. GOOD: Selecting the three limits that directly affect the service’s SLA.
BAD: Saying “I would provision more resources to be safe.” GOOD: Quantifying the cost‑benefit ratio of the additional headroom.
BAD: Accepting the first compensation figure without data. GOOD: Counter‑offering with a market‑based equity percentage and a precise signing‑bonus amount.

FAQ

What should I bring to the capacity planning exercise?

Bring a concise slide that shows demand growth, supply limits, and a risk‑impact table. The slide must fit on one screen and be ready to narrate in under two minutes.

How long does the whole interview loop usually take?

The loop typically spans 14 days from the first phone screen to the final offer, with each round lasting 45 minutes except the capacity exercise, which is 30 minutes.

What is a realistic compensation package for a senior SRE at Amazon?

Base salary ranges from $170,000 to $210,000, equity from 0.05 % to 0.1 % of the company, and a signing bonus between $20,000 and $45,000. Adjust these numbers based on your current total compensation and market data.

Ready to build a real interview prep system?

Get the full PM Interview Prep System →

The book is also available on Amazon Kindle.