# Prepare a leadership product review

Ostrevane · Product-management work sample
All company, customer, partner and incident information is original synthetic material.

## Situation and outcome
Ostrevane is a fictional software company of about 300 people. Its product, Startquern, handles onboarding for employers with many frontline hires: offer letters, document collection, right-to-work checks and payroll sync. It has about 640 paying accounts, mostly with 50 to 499 employees. Its AI document reader, Quillvorn, approves passports and national ID cards without a person; work permits and visas stay in assist mode, where Quillvorn suggests and an HR reviewer approves.

You are the Startquern product manager. The Q3 product review with the leadership team is on Thursday 15 October, 10:00 to 11:00, chaired by the VP Product, who is your sponsor. Today is Friday 9 October, after the 15:00 pre-review of the draft with the VP Product and the functional leads; its feedback is in the correction note in your download. The final pre-read is due to the attendees on Monday 12 October at 17:00. A review-prep agent has drafted the pre-read, and nobody has checked it.

Your outcome is a **final pre-read ready to send**, with the decisions on its first page, and a **note to the VP Product** that says what she must agree before it goes out and what record follows each outcome of the review. A justified "this decision cannot be taken as briefed, and here is what I need first" is a valid result. Agreeing the brief with the VP Product (it is supplied as a record), running the meeting, recording or communicating the outcome, Trust and Compliance's permit review, the security review, settling any customer or sales commitment, and hiring are not part of this work.

The evidence period is 2026-07-01 through 2026-10-09 inclusive. Examine 58 source records from review briefs and operating rules, the July review's record, strategy documents, metrics and finance extracts, engineering notes, a CRM export, win/loss interviews, customer tickets, escalation records and a leadership channel. The records are supplied, not run by you. The AI draft is not evidence authority. Work with the supplied source IDs and context; do not contact fictional organisations or seek private data.

## Your role
You are joining the team as Product Manager.

You own the Startquern pre-read. Choose and reconcile the sources, frame D1 and D2 with options and consequences, and write the note to the VP Product. You may recommend; you may not decide for leadership, change the decisions sought without the VP Product, commit capacity, hires or customer dates, or waive a sign-off.

## About the team
A B2B software company's product team preparing its Q3 product review for the leadership team, chaired by the VP Product.

A review-prep agent drafts the pre-read from the connected tools, a notes assistant summarises meetings and sales reps tag loss reasons in the CRM. People own the figures, the sign-offs, the decisions and the outcome record.

The CEO decides investment, target segment and cross-functional priority after leadership discussion; the VP Product chairs, agrees the pre-read and confirms the outcome record; finance and the data team own the figures; Trust and Compliance signs off any change to what Quillvorn approves without a person; the PM may recommend but decides nothing and makes no customer commitment.

## Work to complete
1. Read planning/review-brief-template.md and sources/coverage-and-access.md before making claims about figures, customers or dates. The review brief, the operating rules and the calendar are among the first four records.
2. Audit the AI draft against the original records, row by row. Change only what the evidence warrants. A row that still stands is kept, with its reason.
3. Report each goal and theme with the figure its owner confirmed, and mark any figure that is not confirmed. Count accounts, rows and opportunities separately: a repeated, copied or relayed record is not a new account, and a tally is not evidence about accounts until it traces to them.
4. Tell apart what a record confirms, what an AI tool wrote, what a person claimed and what a plan assumes. Keep contrary evidence visible. Weigh a serious problem by its harm, not by how few cases there are. Say what is not yet known.
5. Frame each decision so that leadership can answer it at this review: the question, the real options, the consequences of each against the capacity the team has, who decides and what happens after. Check the records on who decides what, and on what must happen before a decision is brought to leadership. If a decision cannot be put as briefed, say so and say what you need and from whom.
6. Read the correction note (updates/U001-pre-review-feedback-v1.md) with the other sources and apply it. Carry each repair through the pre-read, the evidence table, the sponsor note and the revision log, and keep the original wording of a source that the correction disputes.
7. Keep to your role. A recommendation is optional and must be labelled as yours. Leadership decides, and you commit no capacity, hire or customer date.

## Time
You have 90 minutes of work time. The clock starts when you press Start.

After that comes a 10-minute break, then a written interview: up to eight typed questions about your own files, for about 15 minutes. If you submit early, your break starts straight away.

## Mid-task correction
There is no mid-task correction in this take-home. Everything you need is in the files you download at Start, including the correction note **updates/U001-pre-review-feedback-v1.md**. Read it with the other sources, apply it where it changes your work, and cite it as U001.

## Tools and help
Use any tools and help you like: AI tools and agents, search, your own notes, friends or colleagues. You don't need a paid tool, an API key or an account.

Nothing you do in your own tools is watched. Only your files and your interview answers are reviewed. When you submit, you can say which tools or help you used. That is optional, and leaving it empty is never held against you.

## Submission
Hand in one ZIP with these files at its top level. Start from the copies in the templates folder.
- README.md: list every file you hand in and say how the correction note (U001) changed your work, or that it did not.
- pre-read.md: the final pre-read, at most 1,000 words, about two pages at 500 words a page; the word limit is what is checked. Put the decisions on page 1, each with the question, the options, the consequences of each, who decides and what happens after. Then give the Q3 results against goals G1 to G4, the status of every roadmap theme T1 to T4 and the main risks, in plain words for leaders outside product, with every figure marked confirmed or not.
- evidence.csv: the template already has one row for each record, R001–R058, with its record ID and account ID filled in. Fill in the topic column (D1, D2, G1 to G4, T1 to T4, risk or context) and the disposition column (supports, contrary, confirmed figure, unconfirmed figure, status, risk, claim, context or not used are suggested; your own word is fine) on every row. A record that bears on more than one topic gets an extra row for each; never remove a record's row or retype its IDs. When a correction moves a record to another account, leave its account_id exactly as supplied and add a column called corrected_account at the end of the table; fill it only on the moved record's rows (for example, R034 keeps its original account_id and gets corrected_account A001), and say why in the rationale. Add a rationale and an evidence_version where they help. Cite the correction note as U001.
- sponsor-note.md: the note to the VP Product, at most 400 words: what she must agree before the pre-read goes out and who owns each open item, which figures are still unconfirmed, the objections to expect, and what record follows each outcome of the review.
- revision-log.md: material changes to the AI draft, verification checks and remaining gaps, at most 500 words. No private prompts or unrelated files.

Words are counted in the prose of each document you write; headings, tables, the evidence file and the README do not count.

Use plain text files only: .md, .txt, .csv or .json. Put your work in the files themselves; a link to a document somewhere else does not count. Your ZIP can be up to 5 MB, or 20 MB unzipped, with at most 39 files of up to 2 MB each. Programs, macros, password-protected ZIPs and ZIPs inside your ZIP are not accepted.

Upload your ZIP on your take-home page and wait for **Received and checked**. It normally appears within a few seconds of the upload finishing. If it has not appeared after a minute, upload the ZIP again; if it still does not appear, use **Help and support**. You can replace your ZIP until you submit. Submitting is final and starts your break.

If time runs out before you submit, the last ZIP marked **Received and checked** is handed in for you. If none was received, the team is told that you need support.

If something goes wrong, use **Help and support** on your take-home page. A technical problem is never counted against you. Writing to us does not pause the clock.

## What happens next
The written interview covers these parts of the job:
- Checks sources and figures
- Weighs uncertainty and risk
- Checks and repairs the AI draft
- Hands off clearly
- Explains the reasoning
- Frames decisions for leadership

An AI then drafts a review from your files and your answers. Each point quotes what it rests on. A person at the company reads the review and makes any decision. No one is hired, rejected or ranked automatically.
