# Turn approved requirements into stories the team can build

Velquarnith · Product-management work sample
All company, customer and data in this exercise are original synthetic material.

## Situation and outcome
Velquarnith is a fictional software company. Its product, Velquarnith Payables, runs bills, vendor records, and scheduled and batch payments by ACH and wire for small and mid-size businesses. The sponsor, the Head of Payments, has approved one feature, payment approvals: a second person approves a payment at or above a company's threshold before it is sent. The approved requirements (FR-1 to FR-10, QR-1 to QR-3 and an out-of-scope list) were signed off on 2026-09-16 and are settled; do not reopen them. Today is 2026-10-01, and the Payables delivery team's first sprint with this feature starts on 2026-10-05.

An AI agent has already written a first story set from the requirements and the source log, and the delivery team walked it once on 2026-09-29. The draft is not an authority. Your outcome is a story set the delivery team can pick up. Each story is marked ready or not ready against the team's definition of ready (context/story-template.md). You propose readiness for the team to confirm; you do not contact the team. The ready stories are in order, with acceptance criteria a tester could make fail, the team's size, a link to the approved requirement and flags for work other teams must deliver. An open-questions log covers everything else. A well-argued "not ready" set for the whole epic is a valid outcome: each story is still written and marked not ready, tied to the question that blocks it, and each question has an owner and the answer needed.

The story set does not change the approved requirements, plan the sprint or commit capacity, negotiate other teams' dates, replace the delivery team's estimates, build or test anything, or decide the release.

The period is 2026-06-15 through 2026-09-30 inclusive. Work with 61 sources from 27 origins (customers and internal teams): the approved requirements, design and analytics notes, team policies, dependency notes from other teams, an incident review, refinement notes, support tickets and AI-summarised account calls. Use the supplied source IDs and origin codes; do not contact the fictional teams or customers or seek private data or live queries.

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

Take the epic to backlog-ready: stories with acceptance criteria, trace links, the team's estimates, dependency flags and a status for each against the definition of ready, plus an open-questions log for what is not ready. You propose each story's readiness against the definition of ready, and the delivery team confirms it; you do not contact the team, you hand in your proposal for it to confirm. Approved requirements are the sponsor's to change, other teams own their dates, and sprint commitment belongs to planning.

## About the team
A payments software team breaking an approved feature into user stories for its delivery team.

Agents may draft stories and criteria; a person checks them against the requirements, team notes and sources.

People own story readiness with the delivery team, any change to approved requirements, other teams' dates and sprint commitment.

## Work to complete
1. Read sources/coverage-and-access.md and context/story-template.md first. Note what the sources cover and what they cannot show.
2. Audit the AI first story set against the sources. Keep, correct or reject each row and say why. Keep evidence that points the other way visible, and tell apart a measured figure, a person's claim and a gap. Where a figure depends on joining two sources, do the join.
3. Count sources, origins and observations separately. A source that repeats or relays another counts once as an observation. Treat severity separately from frequency. Show each count wherever you state it in stories.md or open-questions.md, with the source and origin IDs behind it. In evidence.csv, the observation_group column is yours to fill: give sources that report the same observation the same short label of your choosing (for example "bank-change-1" on both S028 and S029, which report one incident). A source that holds two separate observations gets a second row with its own label, and an origin with several sources counts once as a customer.
4. Write the stories in the team's format. Each gives a user role something it can use and is sized for the team. Every acceptance criterion carries the values a tester needs. Every approved requirement (FR-1 to FR-10 and QR-1 to QR-3) has a story or a logged reason, and the out-of-scope list stays out. Size is the delivery team's, and you do not estimate: carry the team's number where what a story lets its user do is unchanged (adding values the sources already give to its criteria does not change that). Where you split, merge or add a story, or widen what it does, write "awaiting team estimate" instead of a number. Under the team's definition of ready such a story is not ready until the team sizes it, so list it in the open questions for the delivery team; you may say that everything else in it is ready.
5. Mark each story ready or not ready, put the ready ones in order, and flag what other teams must deliver with the team, the date and the effect if it slips. Where a source gives no date, ask for it in the open questions and still say the effect.
6. Log what is not ready or not settled in the open questions. Each has an owner and the answer needed. Decide only what is yours; name who decides the rest.
7. Read the correction note (U001) with the other sources and apply it. Carry each justified revision through every file. Before-and-after entries in revision-log.md are enough to keep the earlier version visible. The revision log is your own account of how you worked.

## 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-payments-rails-and-crm-correction.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.
- stories.md: the story set, about 1,800 words at most, in the team's format, with the order and a status for each story.
- evidence.csv: the template already has one row for each source, S001–S061, with the source ID and origin code filled in. Do not type IDs. Fill in the story each source bears on and how it is used. Put one story in each row; when a source bears on two stories, add a second row for it. Name your own stories; a use such as trace, supports, contrary, dependency, constraint, context, out of scope or not used is suggested and free text is accepted. Where you state how many customers asked for something, the rows for those customers should show it, and their use column should say they asked (or that the request is out of scope, for one you decline). Record corrections explicitly, and add rationale, observation_group and evidence_version values as useful. Cite the correction note as U001.
- revision-log.md: material changes to the AI first story set, verification checks and remaining gaps, about 500 words at most. No private prompts or unrelated files.
- open-questions.md: what is not ready or not settled, each with an owner and the answer needed, about 500 words at most.

Words are counted in the prose of each document you write, including words in table cells; headings, table borders, source IDs and requirement IDs, 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 counts
- Judges what the evidence supports
- Checks and repairs the draft
- Hands the team ready stories
- Explains their choices
- Writes testable stories

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.
