Research on work and agents

Take-home assignment collection

Thirty original synthetic exercises, shared as material for learning about work with AI.

These are the existing PM assignment packages. Each download contains the candidate-facing brief, synthetic sources and working templates. Corrections are included from the start for self-directed practice. Reviewer guides, live submissions and account records are separate.

Use the synthetic records to examine the work and compare approaches. No purchase, registration or hosted assessment is required.

Make a trustworthy feedback readout

Help a fictional product team turn conflicting customer notes into a trustworthy readout. Review an AI draft, make the evidence and gaps clear, then explain your decisions. This exercise samples synthesis, verification, repair and handoff, not complete PM capability.

Product Management · synthetic exercise · version 1.1.0

Read the brief · Download the exercise · Explore its workflow

Revise a product strategy for leadership review

Help a fictional product team revise an approved strategy for a leadership review. Check an AI-revised draft against a quarter of customer, sales, usage and pilot evidence, change only what the evidence warrants, then explain your decisions. This exercise samples evidence use, judgment, verification, repair, focus and handoff on an existing strategy, not alignment, the leadership decision or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Investigate a drop in checkout conversion

Help a fictional marketplace team find out why checkout conversion fell. Check the data can be trusted, find where the change really sits, audit an AI first analysis and hand a clear findings note to the VP who decides. This exercise samples checking data, splitting a change, judging evidence, repair and handoff, not complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Write a requirements document ready for review

Help a fictional healthcare software team turn stakeholder inputs and an AI first draft into a requirements document that engineering, design and the sponsor can review. Check the draft against the inputs, repair it, write testable requirements, apply the privacy checklist and list the open issues with owners. This exercise samples checking sources, repair, requirements writing and handoff, not complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Build a proposed product roadmap for approval

Help a fictional product team turn planning inputs into a proposed 2027 roadmap for the VP Product's approval review. Check an AI-drafted roadmap against customer, support, partner and capacity records, place each item against what each squad can build and the dates others depend on, and hand the VP the decisions that are theirs. This exercise samples evidence use, judgment, verification, repair, sequencing and handoff on one roadmap, not facilitation, negotiating other teams' dates, the approval itself or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Propose pricing and packaging for a new AI feature

Help a fictional contract-software company decide how to price and package a new AI feature. Work out how it should be sold, check an AI first proposal against the pilot, customer and cost evidence, and hand a validated proposal to the CEO, who decides. This exercise samples value and cost analysis, packaging, checking and repair, and handoff, not complete pricing or PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Build the business case for a monitoring service

Finish and repair a business case for a fictional equipment maker's monitoring service before the Investment Committee meets. Build the options and yearly figures to finance's rules, audit an AI first business case, take in the assumption review and hand the committee a proposal with its conditions, assumptions and owners. This exercise samples building and repairing a business case, judging evidence and handoff, not complete PM capability: the committee's decision, the stakeholder meetings, pricing design and the forecast stay outside it.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Plan a launch and report what is ready

Plan the launch of a new plan for a fictional running app and report what is ready. Check every status against the team's records, audit an AI first draft of the plan, set out the decisions that belong to the sponsor and revise when the release manager replies. This exercise samples building a launch plan and checklist, status from evidence, judgment about risk, repair and handoff, not complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Propose a north star the team can trust

Help a fictional ordering marketplace propose a north star, inputs and guardrails for one product area. Judge the candidate metrics against the evidence, audit an AI first draft, define each metric so it can be computed only one way and hand a feasibility review to the leader who adopts it. This exercise samples framing value, judging evidence, metric definition, repair and handoff, not complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Rollouts competitor analysis

Build a short competitor analysis for a fictional developer-tools product from sales, customer and public records, audit an AI first draft of it, and name who decides each response. Then keep it current: a monitoring check brings new competitor moves and two owners' replies, and you judge which changes are material and revise or log no change. This exercise samples scoping alternatives, weighing evidence and confidence, proposing responses to the people who decide them and keeping an analysis current. It does not sample real stakeholder review, repeated monitoring over quarters or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Conclude an A/B test and close it down

Conclude a finished A/B test for a fictional language-learning app against the rules set before launch. Check the AI first draft of the readout against the team's records, record a verdict, report it plainly to the growth review, write the request to close the experiment down (nothing is carried out) and revise when billing and lifecycle messaging reply. This exercise samples concluding a test, reading it by its plan, judgment about harm and uncertainty, repair and handoff, not designing a test or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Propose a priority order for the next quarter

Help a fictional product squad turn scorecards, votes and a positions thread into a proposed priority order for next quarter's planning review. Check an AI-drafted order against the records, draw the cut line at the capacity the team really has, keep every dependency in place, and show who agrees, who objects and who decides. This exercise samples evidence use, judgment, verification, repair, ordering against capacity and handoff on one planning period, not facilitating the review, negotiating other teams' dates, the decision itself or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Decide a model change on evidence

Check the evidence behind a proposed model change to a fictional business software assistant and recommend, for the decision owner, whether to adopt it. Audit an AI-drafted readout against the test run, the human ratings and the sample answers, build the scorecard and suite plan the decision should rest on, and revisit your work if a scheduled update affects it. This exercise samples reading evaluation evidence, weighing severity, observed results and uncertainty, repairing an AI draft, designing criteria and thresholds and handing a decision to its owner. It does not sample running models, graders or raters, tuning, release, working with the ML team or the Director over time, or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Write the opportunity brief for medication refills

Help a fictional pet-care app decide whether medication refills are an opportunity worth a closer look. Check an AI first draft against mixed evidence, write the brief the Head of Product reviews, and revise it when a feasibility and data check-in changes the facts. This exercise samples framing a problem, judging evidence, repair and handoff. It does not sample running validation research, building the business case or winning the review, and it is not complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Get a product change through specialist review

Take a fictional fintech's AI card-dispute feature through legal, privacy, security, accessibility and AI-risk review and report honestly where it stands. Check an AI-written intake summary against the records, route the change to every review it needs, keep each reviewer's own findings and severity, and give the initiative owner a status they can act on before the launch meeting. This exercise samples evidence use, judgment, verification, repair, routing and handoff on one change, not the specialist determinations, the partner bank's decision, the build of any fix or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Prepare a leadership product review

Prepare a fictional B2B software company's quarterly product review for its leadership team. Check an AI-drafted pre-read against the records, report the quarter's results with the figures their owners confirmed, frame two decisions so leadership can answer them on the day, and write the sponsor the note that says what she must agree before the pre-read goes out. This exercise samples evidence use, judgment, verification, repair, framing decisions and handoff on one review cycle, not agreeing the brief, running the meeting, pre-briefing leaders, recording or communicating the outcome, or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Size and segment a market and recommend target segments

Size and segment a market for a fictional scheduling-software company and recommend which new segments to pursue. Check an AI first sizing against the sources, build each segment from who does the job and what they pay, check the company's bar and revise when the research unit replies. This exercise samples sizing, segmentation, verification and repair, and handoff to the decision owner, not the decision, personas, positioning, pricing or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Recommend how to source invoice capture

Recommend how a fictional accounts-payable software company should source invoice capture, for its chief product officer's review. Check the AI first draft against the team's records, compare building, buying and partnering on the same terms, assess dependency risk and the way out, propose a way forward marked proposed, leave the decision with its owners and take account of the vendor, counsel and leadership notes of the final week. This exercise samples sourcing judgment, repair and handoff, not vendor selection, negotiation, legal review or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Prepare an industry analyst briefing

Prepare a briefing deck and written answers for an industry analyst's questionnaire at a fictional developer-platform company, and hand the pack to clearance. Check an AI first draft against the sources, say only what is true and cleared, defer the rest with honest holding lines, apply the clearance pre-check note and tell the reviewer what to decide. This exercise samples building and verifying the briefing pack up to clearance, not the reviewer's decision, delivering the briefing, building the relationship, follow-ups, public talks or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Turn approved requirements into stories the team can build

Help a fictional payments software team turn an approved feature and an AI first story set into stories the delivery team can pick up. Check the draft against the approved requirements, team notes and customer sources, slice the work so each story gives a user something usable, write acceptance criteria a tester can fail, flag what other teams must deliver, and say plainly which stories are ready and which are not. This exercise samples checking sources, repair, story writing and handoff to a delivery team, not facilitating a refinement meeting, estimating, planning a sprint or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Recommend whether a release is ready to ship

Take over the go/no-go preparation for a fictional support-software company's October release and tell its Head of Product whether the release candidate is ready. Check an AI-written first recommendation against the records, work out which items can ship and which should be held back, say how any held item is withheld and what must be verified, and revise when the release manager's reply changes a fact. This exercise samples evidence use, judgment, verification, repair, release scoping and handoff on one release candidate, not the verdict itself, the deployment, the fixing of defects, negotiation with customers or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Run the monthly indicator review

Run one month's indicator review for a fictional maker of warehouse trucks and its Kelvorwick app: send the agenda, check the snapshot, repair an AI-drafted summary, class every indicator, get each material shift an owner who has really accepted it or an open item with its route, and publish the summary. This exercise samples evidence use, judgment, verification and repair, handoff and the call on what matters and who owns it on one review cycle, not agreeing the indicator set, investigating causes, holding the meeting, following actions through over time, or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Q3 trend brief for product leadership

Produce one edition of a quarterly market, technology and regulation trend brief at a fictional remote patient monitoring company, for its leadership review. Check an AI first draft and the scanned signals against the sources, assess each item (classification, horizon, relevance to strategy, uncertainty), record for each regulatory statement whether counsel is needed or it is factual only, propose follow-ups with owners, route anything urgent now, and revise when the final payer rule and a CRM fix change the record. This exercise samples producing and verifying the edition up to the review, not the review itself, accepting follow-ups, publishing, counsel's interpretation, scheduling the product's response to a rule, or complete PM capability.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Segment positioning proposal

Write the positioning and value proposition for three target segments of a fictional maker of compressed-air monitoring for manufacturing plants: audit an AI-drafted set against the customer, sales and market records, decide per segment whether the evidence is enough to position on, write each proposition and statement or a gap record, propose proof points with the sign-off each needs, prepare the review for the two approvers, then revise when the message-check findings arrive. This exercise samples evidence use, judgment, verification and repair, handoff and the call on what each segment's evidence supports, on one proposal. It does not cover choosing segments, pricing, campaign copy, running message tests, the approval decision, influence over time, leadership through people, commercial impact or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Plan a safe staged switch to a new alert engine

Plan the staged release of a replacement alert engine for a fictional maker of maintenance monitoring software: check an AI-drafted plan against the records, work out what the test results cover, design the stages with the customers who move in each, their entry and exit criteria, rollback triggers and roster steps, check every date against what the records allow, and put the decisions that only the approvers can take in front of them. This exercise samples evidence use, judgment, verification and repair, handoff and staged rollout design on one release, not running a beta, deciding whether a stage may widen, measuring results, running customer migrations, executing the old engine's retirement, influence or negotiation over time, or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Get price-alert tracking ready to trust before launch

Get a new price alerts feature's tracking ready to trust before launch for a fictional resale marketplace: check an AI-drafted readiness summary against the sources, verify each event and metric against the team's own checklist, write the final instrumentation specification, route what fails to the people who own it, and apply the privacy and engineering replies. This exercise samples evidence use, judgment, verification and repair, handoff and the design of a specification on the repair and verify end of one feature's instrumentation, not choosing the metrics, building the events, measuring the launch, investigating a metric move, the formal privacy intake, following fixes through over time, or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Prepare the portfolio review pre-read

Prepare the pre-read for a half-yearly product portfolio review at a fictional software company: check an AI-drafted options analysis against the records, put six products on one basis, test each submission's case, carry forward the last cycle's conditions, frame funding options that fit one engineering budget, and say which decisions are not ready. This exercise samples evidence use, judgment, verification and repair, handoff and portfolio trade-offs on one review cycle, not setting the brief, running the committee session, pre-briefing or negotiating with executives, recording and tracking rulings over time, or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

MVP scope for first customers

Define the smallest version of a new warehouse phone app that can prove its value with the first customers, for a fictional order-software company: check an AI-drafted MVP definition against the records, decide which capabilities fit the build capacity and which are left out and why, set a target with a baseline and a method for each first customer, recommend approval with conditions or not ready for the decision owner, and apply the engineering and data replies in the correction note. This exercise samples evidence use, judgment, verification and repair, handoff and scope definition on the define-and-propose end of one MVP, not recruiting the first customers, building the product, the build gate decision, pricing and launch, writing requirements, or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Triage two weeks of feature requests

Work through two weeks of a fictional hosted build platform's feature requests and feedback: decide for each item where it goes, whether it duplicates something already shipped, declined or on the backlog, what the person behind it needs, and what to tell them. Check an AI-written first pass against the records, count demand by account rather than by message, send what needs someone else now to the right owner, and keep every reply free of delivery dates and of commitments about delivery, customer-specific builds or contracts. This exercise samples evidence use, judgment, verification, repair, routing and replies, and handoff on two weeks of items. It does not show prioritising the backlog, deciding a customer's contract request, diagnosing bugs or security incidents, following up with requesters over time or complete PM capability. The written interview is part of this take-home, because it is where you show the reasoning behind your written work.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

Answer the product questions in a tender

Help a fictional remote-monitoring supplier give a public-sector buyer tender answers it can stand behind: answer the product-owned questions and hand them to the bid owner. Check an AI-drafted answer set against the released product and the verification on file, route security, privacy and legal statements to the people who own them, keep every answer something the company can stand behind, apply the issuer's clarifications and tell the bid owner what is ready, what departs from the requirements and what is theirs to decide. This exercise samples checking and assembling product answers up to the bid owner's clearance, not the bid owner's clearance or submission, specialist determinations, pricing, negotiation after award or complete PM capability. The written interview is optional; it is a good place to ask about the reasoning behind any answer.

Product Management · synthetic exercise · version 1.0.0

Read the brief · Download the exercise · Explore its workflow

These are synthetic learning exercises. Third-party source material remains subject to its original terms.