Aug 24–31 interview digest: Google's nine-week team match, Riot's eight-stage loop, and a $1.2M Staff offer (2026)

The Aug. 24–31, 2026 window produced a stronger preparation signal than a pass-rate signal. A Google candidate describes a nine-week path from onsite feedback through team matching, a Riot Games candidate has passed three rounds and faces five more interviews, and an Anthropic post reports a question about what happens if company stock falls to zero. The compensation side has cleaner numbers: a Google L5 offer lists a $247K base, $900K of equity over four years, a $75K sign-on payment, and a 15% target bonus; a Databricks Staff offer lists $1.2M in total compensation.

Blind pages display relative ages rather than absolute native publication dates. The article preserves those ages and labels the timestamp gap. Reddit supplies two detail pages with absolute creation times in the window. Glassdoor's Google interview page is an aggregate/index surface, while current Glassdoor report pages did not yield a usable current-week report body or report-level engagement. 1 The available material contains zero company-wide pass-rate denominators. The issue therefore uses recent accessible reports and clearly marked discussion markers instead of claiming a cross-platform vote ranking.

The week in one read

Three details change the drills worth running before an interview.

A successful onsite can still lead to weeks of team matching and silence. One Google L4 candidate reports positive onsite feedback within a week, two unsuccessful team matches in weeks 1–3, silence in weeks 4–6, a third match call in week 7, and human-resources and compensation approvals in weeks 8–9. 2
A loop may extend well past the familiar coding, technical, and behavioral labels. The Riot Games candidate says three rounds are complete and five additional interviews remain for a software-engineering role. 3
A headline TC figure needs its ingredients and timing. The largest disclosed package this week is a Databricks Staff offer with $260K base, $900K equity, and $30K bonus. The post compares it with a seed-stage AI startup offer and attributes the culture claims to people involved in those offers. 4

Evidence at a glance

This week's evidence set

Current-window detail pages and the fields they expose.

First-person detail pages used	11
Concrete prompt or format signals	4
Company pass-rate denominators	0

Loading stats card…

The 11 detail pages include nine accessible Blind posts and two Reddit posts. Blind shows 2d, 3d, 4d, 6d, or 7d on the selected pages, with no absolute native publication date. Reddit gives Aug. 30 and Aug. 31 creation times for the two posts used below. Poll participants, comment totals, and Blind's generated comment summaries describe discussion activity; they do not create an interview denominator.

Report cards

Company or scope	Freshness shown	Round, task, or process	Reported outcome	Engagement and evidence boundary
Google	`4d` on Blind; absolute native date unavailable	Onsite passed in early June; two team-match calls in weeks 1–3; silence in weeks 4–6; a third match call in week 7; human-resources, packet, and compensation approvals in weeks 8–9.	Signed at $230K new TC, up from $195K former TC.	The page's AI summary covers 10 comments and was generated Aug. 27. The summary is discussion metadata; the first-person timeline supplies the process facts. 2
Anthropic	`7d` on Blind; absolute native date unavailable	The author says an interviewer asked what would happen if the company's stock dropped to zero. Role, level, round, and interviewer context are undisclosed.	Outcome undisclosed.	The page displays a generated summary based on 181 comments, dated Aug. 29. The summary is labeled unverified, so the first-person sentence is the usable evidence. 5
Riot Games	Posted Aug. 31 at 1:18 a.m. UTC-05:00	Three completed rounds: one coding, one technical, and one behavioral; five additional interviews remain for a SWE role outside game development.	Passed the first three rounds; final outcome undisclosed.	Reddit reports a score of 7, an 81.8% upvote ratio, and 11 comments. The Reddit comments route was unavailable, so community reaction cannot be summarized. 3
Salesforce	`3d` on Blind; absolute native date unavailable	The author says final rounds for an SMTS role happened two weeks earlier; the recruiter was waiting for feedback and stopped replying after the candidate mentioned competing offers.	Decision pending in the post.	The page exposes a generated 10-comment summary dated Aug. 29 and no usable compensation number. The post supports a communication-risk signal, not a hiring-rate claim. 6
Staff/Principal hiring	`3d` on Blind; absolute native date unavailable	A seven-round Staff/Principal Software Engineer loop; the recruiter called the overall feedback positive, while the team declined to proceed.	Rejected; other opportunities were mentioned.	The page includes a 17-participant poll and a generated 10-comment summary dated Aug. 29. Those figures mark discussion volume, not a pass rate. 7
Google L5	`3d` on Blind; absolute native date unavailable	Offer evaluation rather than an interview report.	Offer received; the author says one negotiation round has already happened.	Displayed components: $247K base, $900K equity over four years, $75K sign-on, and 15% target annual bonus. The page has 10 comments and a generated summary dated Aug. 29. 8
Apple and Anduril	Posted Aug. 30 at 9:04 p.m. UTC-05:00	Offer comparison for a candidate with two years of DevOps/SRE experience. Anduril's role is backend engineering for internal IT automation; Apple's role is DevOps/SRE work supporting device-manufacturing applications.	Both offers are described as accepted or received at different points; the candidate is deciding whether to switch to Apple.	Anduril's disclosed package is $165K base plus $165K RSUs over four years at IC3. Apple's components are withheld by the post, which says pay and level are similar. Reddit comments were unavailable. 9

Blind's 2d, 3d, 4d, 6d, and 7d markers show relative age at retrieval. The native pages leave the calendar publication date open, so the markers stay in their original form.

The question bank you can practice now

Reported material	Round or format label	Immediate drill	Evidence boundary
An Anthropic interviewer reportedly asked, "what if stock drops to 0"? 5	Round and role undisclosed	Prepare a two-minute answer that separates mission, compensation risk, dilution, liquidity, and the conditions under which you would leave. Then ask the interviewer what the question is intended to test.	One first-person account supports a culture or values-question signal. The source gives no basis for calling it universal.
Riot's author reports one coding, one technical, and one behavioral round, followed by five additional interviews. 3	Three completed rounds; five further interviews; SWE role	Practice a full-loop handoff: solve a coding problem, explain a technical design or implementation choice, answer a behavioral story, and keep one project deep dive ready. Ask what each remaining interview measures.	The post gives formats and sequence, while the five remaining interview topics are undisclosed.
The Google candidate reports two team matches that went to other candidates, a third match, and later packet and compensation approval. 2	Team matching after onsite; timing varies by candidate	Ask whether team matching begins before or after hiring-committee review, how long feedback remains active, how many teams may review the profile, and who owns updates during silence.	One process account gives a timeline. The account does not establish a standard Google timeline.
The senior rejection post describes a seven-round loop with positive overall feedback and a team-level rejection. 7	Seven-round Staff/Principal loop	After each round, write the evidence that supports scope, influence, technical judgment, and team fit. A positive signal in one dimension cannot stand in for the complete senior-level packet.	The post leaves every question and rubric undisclosed.

The preparation change is simple: ask for a written round map before spending a week on a generic problem list. Record the format, duration, interviewer focus, coding environment, feedback timing, and the stage owner. Those fields tell you whether the next session should emphasize recall, implementation, system design, or leadership evidence.

Outcomes and the pass-rate read

The Riot post reports a positive intermediate outcome: three rounds passed and five interviews still ahead. The Google L4 post reports a signed offer after team matching. The Staff/Principal post reports rejection after a positive overall feedback statement. The Salesforce post reports a pending decision after recruiter silence. These are four different endpoints, so the article keeps them in separate rows.

The current set contains no company-wide numeric pass-rate denominator. The six-final-round, zero-offer story from the prior window remains outside this week's evidence and is excluded here. This week's pages describe individual outcomes, active processes, or offer decisions; they do not connect the same company, level, stage, and outcome across enough candidates to calculate a company rate.

Reddit's Riot detail page reports 7 points, an 81.8% upvote ratio, and 11 comments. The Apple/Anduril page reports 1 point, a 57.1% upvote ratio, and four comments. Blind displays participant counts such as 181 on the Anthropic question, 17 on the senior rejection poll, 82 on the Google/LinkedIn/Plaid poll, and 45 on the Databricks comparison. These are engagement markers. They help locate discussion, while the first-person report determines the interview or offer claim.

Compensation: signed numbers first, benchmarks second

A first-person offer answers, "What did this candidate receive?" An aggregate benchmark answers, "What do reported packages look like at this level?" Negotiation gets clearer when those answers occupy separate rows.

First-person figures from this week

Company or comparison	Level and location	Base	Equity	Bonus	Sign-on	TC and status
Google	L5; Bay Area	$247K	$900K / 4 years	15% target	$75K	The page lists an offer and says one negotiation round has already happened. The page does not publish its own TC. A simple annualization gives $225K equity plus a $37.05K target bonus, or about $509K before sign-on, if the equity is spread evenly. 8
Google	L4; location undisclosed	Undisclosed	Undisclosed	Undisclosed	Undisclosed	$230K new TC, versus $195K former TC; signed without negotiation. 2
Databricks	Staff Software Engineer; San Francisco	$260K	$900K	$30K	Undisclosed	$1.2M TC on the offer card. The post compares the package with a seed-stage AI startup offer. 4
AI startup	Staff Software Engineer; San Francisco	$325K	$600K	Undisclosed	$30K	$925K TC on the same offer comparison. The post says the team appeared to have normal hours; that culture description belongs to the anonymous post. 4
LinkedIn	Software Engineer; Sunnyvale; Entry on the offer card	$146K	$44K	$15K	$10K	$204K recurring TC in the candidate's comparison. 10
Apple	Backend Engineer; Austin; ICT2	$128K	$24K	None shown	$15K	$152K recurring TC in the candidate's comparison. 10
Anduril	IC3; Los Angeles	$165K	$165K / 4 years	Undisclosed	Undisclosed	Offer described by a Reddit author with two years of DevOps/SRE experience. Apple pay and level are described only as similar. 9
Google	L4; Mountain View	$195K	$55K	$20K	Undisclosed	$270K TC on a Google/LinkedIn/Plaid offer card. 11
LinkedIn	IC2; Sunnyvale	$180K	$70K	$18K	$20K	$268K TC on the same offer card. 11
Plaid	E4; San Francisco	$200K	$120K	Undisclosed	Undisclosed	$320K TC on the same offer card. 11

The Google L5 arithmetic is a comparison aid rather than a source-published total. Ask whether the equity figure is an annual grant or a four-year grant, whether the target bonus is paid on base salary, when the sign-on payment arrives, and whether a refresh grant is part of the written offer.

Fresh aggregate annual TC anchors

Levels.fyi pages for Google, Meta, Amazon, and Stripe were each marked Last updated: 8/31/2026. The figures below are aggregate software-engineer submissions and belong beside a signed offer only as level anchors.

Company	Mid-to-senior software-engineer levels	Aggregate annual TC anchors	Page context
Google	L4–L7	L4 $292K · L5 $428K · L6 $689K · L7 $941K	The page displays several possible equity vesting schedules, including 38/32/20/10, 33/33/22/12, 50/28/12/10, 36/28/20/16, and 25/25/25/25. These are page examples, not a universal employee term. 12
Meta	E4–E7	E4 $296K · E5 $427K · E6 $685K · E7 $1.32M	The page context shows 25%/25%/25%/25% RSU vesting. 13
Amazon	L5–L7	L5 $274K · L6 $388K · L7 $685K	The page context shows backloaded RSU vesting at 5% / 15% / 40% / 40%. 14
Stripe	L3–L5	L3 $434K · L4 $683K · L5 $710K	The page shows one-year 100%, two-year 50/50, and four-year 25%/25%/25%/25% examples, with refresh availability after nine months. 15

The benchmark table gives a negotiation range by level. The first-person table gives terms to verify line by line. Amazon's page context makes year-one cash especially important when equity is backloaded. Stripe's page context makes the vesting horizon and refresh timing worth asking about. Google and Meta show several possible schedules, so the written grant matters more than the company label alone.

Company read-through

Company	Current-week signal	What a candidate can use	Evidence boundary
Google	A signed L4 offer followed a nine-week team-match path; a separate L5 offer exposes unusually complete components. 2 8	Prepare for post-onsite waiting and ask how team matching, packet approval, and compensation approval connect. Use the L5 components as a checklist for negotiation.	Two anonymous offer/process posts; no company pass rate.
Anthropic	One first-person post reports a stock-at-zero question. 5	Prepare a values answer that includes compensation and downside risk.	Round, role, level, outcome, and exact interviewer context are undisclosed.
Amazon	The fresh aggregate page gives L5–L7 TC anchors. 14	Compare a written Amazon offer with level, vesting, cash timing, and location.	The current-week set contains no Amazon interview-question detail.
Meta	The fresh aggregate page gives E4–E7 TC anchors. 13	Use the level mapping when another offer uses a different title or ladder.	Benchmark evidence only; no current-week first-person Meta report.
Stripe	The fresh aggregate page gives L3–L5 TC anchors and vesting examples. 15	Ask how grant duration and refresh timing affect annual and year-one value.	Benchmark evidence only; no current-week first-person Stripe offer.
Salesforce	One SMTS candidate reports two weeks of silence after final rounds and a competing-offer disclosure. 6	Set a follow-up cadence and ask who owns the decision when recruiter communication stops.	One pending process; no question text, outcome, or company rate.
Apple / LinkedIn / Plaid	Current offer comparisons expose large differences in recurring TC and location, with role and level labels attached. 10 11	Compare recurring pay, location, role fit, network value, and equity timing as separate decisions.	Anonymous offer cards; candidate-specific teams and future outcomes are unknown.

The connected pattern is process uncertainty plus compensation asymmetry. The Google timeline puts waiting after a successful onsite into view. The senior rejection shows that positive feedback can coexist with a team-level decision. The offer cards show why the same headline TC can carry different cash, equity, location, and risk.

Before your next interview or negotiation

Put four sheets beside your calendar: a round map, an unfamiliar-question drill, a post-solution review, and a component-level offer ledger.

Ask for the round map. Get the format, duration, interviewer focus, coding environment, feedback timing, and decision owner in writing. Riot's five remaining interviews and Google's team-match timeline both make the missing process fields expensive.
Practice one unfamiliar question each day. For a values question such as Anthropic's stock scenario, answer the trade-off instead of reciting a mission statement. State how you weigh mission, cash, equity, liquidity, and downside risk.
Use the last ten minutes of every coding session. State complexity, test edge cases, explain trade-offs, and describe one production change. A solved problem and a complete interview answer are separate tasks.
Track senior-level evidence by dimension. For Staff or Principal loops, keep separate examples for technical scope, influence, judgment, execution, and team fit. Positive feedback in one area gives you a signal; it does not replace the full packet.
Build the offer ledger. Record base, target bonus, equity grant, vesting calendar, sign-on payment, location, return-to-office rules, level, and refresh policy. Compare first-year cash, recurring annual value, and four-year value in separate rows.
Set a communication deadline. After a final round, ask when the next update is due and which person owns the decision. If the date passes, send a short written follow-up with the competing-offer deadline and the exact information you need.
Use aggregate benchmarks as anchors. Compare your written package with the matching level on Levels.fyi, then ask the recruiter to explain differences in grant size, vesting, bonus target, sign-on timing, and location. Keep the benchmark separate from the terms you have actually been offered.

This week's pages give four usable preparation signals: a values question, an eight-interview Riot sequence, a nine-week Google team-match path, and a seven-round senior rejection after positive feedback. The offer pages add component-level figures from Google, Databricks, Apple, LinkedIn, Plaid, and Anduril, while the fresh benchmark pages supply level anchors for Google, Meta, Amazon, and Stripe. Each number and outcome stays attached to the source and evidence class that produced it.