Prepare a decision
Decide your next move with more clarity.
Choose a decision. Add your context. ImpactOS measures your readiness and gives you the next action.
No sign up. Nothing is sent by default.
What decision are you preparing?
Everything after this step adapts to the decision you pick, so nothing irrelevant is asked.
One decision at a time.
See the full analysis
Optional sections. Open only what is useful for your current decision.
Full command centerEvery area at once, including the ones this decision does not need.
Your command center
One view over your career, your evidence, your money and your execution. Areas you have not measured stay empty rather than being guessed.
Career readiness
Not measurable yet
No document has been read yet.
Analyse a document to measure it.
Evidence strength
Not measurable yet
No evidence has been read yet.
Publish one artefact a reader can open in under a minute.
Financial readiness
Not measurable yet
No figure has been entered in this session.
Enter a few figures to measure your base.
Execution consistency
Not measurable yet
No weekly history yet.
Produce one proof this week, however small.
Overall decision readiness
Not measurable yet
At least two areas are needed before an overall view means anything.
Open the decision engine and pick the decision you are facing.
Full diagnosticThe longer guided questionnaire and its detailed result.
Run your own diagnostic, no account needed
Pick a path, answer a short set of fixed questions and get a personalised readiness view straight away. This is a structured decision model, not an AI system. Everything stays in your browser.
1. Choose your path
2. Answer the questions
Decide which role to aim for and what proof is missing.
1.Target clarity
How precisely can you name the role, level and type of team you are aiming for?
2.Existing proof strength
What can a reviewer open today that shows you can already do the work?
3.Skill gap severity
How large is the distance between what you can do now and what the target demands?
4.Network leverage
How many people already doing this work would review or refer your output?
5.Weekly time capacity
How many protected hours a week can you actually spend on this?
6.Execution consistency
Over the last three months, how reliably did you finish what you started?
7.Market alignment
How well does what you are building match what target postings actually ask for?
Add your contextNotes kept in this browser only, never uploaded.
Add your context
This beta helps you structure your context. Deep document analysis will be enabled progressively.
Your document
Your file stays on your device. Nothing is uploaded and nothing is stored.
Document type
Accepted formats: .pdf, .docx, .txt, .md, .csv, .xlsx, .json, .yaml, .yml
This text stays in your browser. It is never saved and never sent.
Your context summary
Computed in your browser, with no analysis service.
Describe your situation above to see your context summary here.
Proof portfolioArtefact catalogue and a copyable brief once a diagnostic exists.
Proof Portfolio
BetaA score is only a snapshot. Proof is what changes how other people judge you. Turn each diagnostic and each sprint into one artefact someone else can open.
No proof brief yet
The Proof Portfolio becomes personalised after a diagnostic.
Run the diagnostic first. The artefacts below stay visible either way, so you can see what strong evidence looks like before you start.
Build my Proof PortfolioCase study
- What it proves:
- That you have already moved a real situation from a starting point to an observable outcome.
- Best for:
- Career, freelance consultant and branding paths
- Expected output:
- One page with context, constraint, decision, action and measured outcome.
- First action:
- Write the outcome sentence first, then work backwards to the decision that produced it.
Product brief
- What it proves:
- That you can frame a problem, choose a scope and state what you will not build.
- Best for:
- AI Product Manager and Data Product Manager templates
- Expected output:
- A short brief with problem, user, success criteria, non goals and open questions.
- First action:
- Write the non goals section before the feature list.
Market test
- What it proves:
- That demand was observed rather than assumed.
- Best for:
- Founder SaaS and side project paths
- Expected output:
- A test record with audience, offer, channel, sample size, result and stop criteria.
- First action:
- Write the stop criteria before running anything.
AI evaluation plan
- What it proves:
- That you can judge quality of an AI feature instead of trusting a demo.
- Best for:
- AI Product Manager and career switch to AI
- Expected output:
- A plan with tasks, test set, failure modes, thresholds and who reviews the results.
- First action:
- List the three failure modes that would be unacceptable, then design a check for each.
Data metric definition
- What it proves:
- That you can make a number trustworthy and owned.
- Best for:
- Data Product Manager and analytics oriented roles
- Expected output:
- A definition with owner, source, formula, refresh rule, known limits and consumers.
- First action:
- Name one metric people argue about and write its single definition.
Founder validation memo
- What it proves:
- That an idea was tested against the assumption most likely to kill it.
- Best for:
- Founder SaaS and idea validation
- Expected output:
- A memo with riskiest assumption, test run, evidence collected and the kill, pivot, test or build call.
- First action:
- State the assumption in one sentence and the evidence that would disprove it.
LinkedIn proof post
- What it proves:
- That the work exists in public and can be found by the people who hire or buy.
- Best for:
- Branding, freelance and job search paths
- Expected output:
- A short public post that links to one artefact and states the outcome, not the ambition.
- First action:
- Publish the artefact first, then write the post around what it shows.
Portfolio project page
- What it proves:
- That someone can review your judgement without a meeting.
- Best for:
- Every path once at least one artefact exists
- Expected output:
- One page per project with the problem, the decision, the artefacts and the result.
- First action:
- Create the page with a single project, even an incomplete one, and keep the same structure.
Path templatesStart from a ready made path and jump straight into the diagnostic.
Path templates
What good evidence looks like for five common directions.
AI Product Manager
TemplateBe credible as the person who decides what an AI feature should do, what it must not do and how it is evaluated.
- Required proof
- One written product spec for an AI feature with explicit failure modes
- One evaluation plan showing how quality would be measured
- One shipped or prototyped feature a reviewer can open
- Typical gaps
- Talking about models instead of user outcomes
- No evaluation or guardrail thinking
- No artefact that shows judgement under uncertainty
- Risk signals
- Every claim is a course certificate rather than a decision you made
- No opinion on what should not be automated
- Cannot describe a case where the model was wrong and what you did
Seven day sprint focus: Publish one AI feature spec with success criteria, failure modes and an evaluation plan.
Recommended Decision Lens: The sceptical operator
Data Product Manager
TemplateBe trusted to own a data product where correctness, lineage and adoption matter more than volume.
- Required proof
- One data contract or metric definition document
- One before and after example of a decision improved by better data
- One quality dashboard or quality rule set you designed
- Typical gaps
- No clear metric ownership story
- Confusing reporting work with product work
- No evidence of stakeholder alignment on definitions
- Risk signals
- Cannot name who consumes the data and what breaks when it is wrong
- Quality is described as a tooling problem only
- No adoption measure for anything you shipped
Seven day sprint focus: Write one metric definition with owner, source, refresh rule and known limits.
Recommended Decision Lens: The accountable owner
Founder SaaS
TemplateDecide whether a SaaS idea deserves the next quarter, using demand evidence rather than enthusiasm.
- Required proof
- Ten scripted conversations with people who have the problem
- One live page with a real pricing question and observed responses
- One manual delivery of the outcome for a single user
- Typical gaps
- Audience described too broadly to reach
- No stop criteria written before testing
- Building before the outcome was delivered manually once
- Risk signals
- All evidence comes from people who like you
- No answer to how the first ten users would find it
- The problem cannot be expressed as a frequency and a cost
Seven day sprint focus: Run the cheapest test that could prove the idea wrong this week.
Recommended Decision Lens: The cost of being wrong
Freelance consultant
TemplateBe findable and verifiable for one specific problem, so qualified enquiries arrive without cold outreach.
- Required proof
- One public claim expressed as a problem you solve
- One case write up with context, action and observable outcome
- One reusable artefact a prospect can open before talking to you
- Typical gaps
- A generalist claim nobody searches for
- Claims with no attached artefact
- No repeatable way of reaching the audience
- Risk signals
- Pipeline depends entirely on one referral source
- Cannot state a price with a reason behind it
- Nothing published in the language the buyer uses
Seven day sprint focus: Reduce the profile to one claim and one proof that supports it.
Recommended Decision Lens: The impatient buyer
Career switch to AI
TemplateMove into an AI adjacent role by transferring existing strengths instead of restarting from zero.
- Required proof
- One artefact that applies AI to the domain you already know
- One written translation of your current experience into the target role language
- One conversation with someone doing the target role today
- Typical gaps
- Learning without producing anything openable
- No bridge between the previous domain and the target role
- Target role named too vaguely to prepare for
- Risk signals
- Runway is shorter than the time the switch usually needs
- Every proof is a tutorial reproduction
- No one in the target field has reviewed the work
Seven day sprint focus: Ship one artefact that combines the domain you already know with the direction you want.
Recommended Decision Lens: The hiring manager with ten minutes
Decision historySaved results and how your readiness moves over time.
Example scenarios, synthetic personasRead only reference examples. Both personas are fictional.
Career Impact Map
Maya, 29, business analyst moving toward AI Product Management
Synthetic persona. No real person, employer or project is described.
Goal: Become credible for AI Product Manager roles within six months
Career Readiness Score
55
of 100, Needs more proof
Evidence Strength Score
42
of 100, Needs more proof
Execution Readiness Score
40
of 100, Needs more proof
Why this score? Career readiness
- Target clarity4 of 5, weight 3Narrow the target to one role family and one industry.
- Proof strength2 of 5, weight 3Publish one observable artefact a reviewer can open.
- Skill gap severity2 of 5, weight 2Close the evaluation and failure analysis gap first.
- Network leverage2 of 5, weight 2Add two practitioners who can critique the work in public.
- Time capacity2 of 5, weight 2Protect a fixed weekly block instead of spare time.
- Execution consistency3 of 5, weight 3Ship something small every week rather than one large plan.
- Market alignment4 of 5, weight 2Match the proof to what the target postings actually ask for.
What would improve this score
- Raise skill gap severity: Close the evaluation and failure analysis gap first.
- Raise network leverage: Add two practitioners who can critique the work in public.
- Raise time capacity: Protect a fixed weekly block instead of spare time.
Limits
- Inputs are self reported, so the score reflects what you declared, not verified facts.
- This is decision support, not career, legal, financial, psychological, medical or investment advice.
- A high score does not guarantee an outcome. It means the evidence you hold is more consistent.
Why this score? Evidence strength
- Public artefacts1 of 5, weight 3One published case study would move this the most.
- Measured outcomes2 of 5, weight 3Attach a number and a method to one shipped result.
- Third party validation2 of 5, weight 2Get a practitioner critique on the record.
- Recency of evidence4 of 5, weight 2Keep the newest proof less than three months old.
What would improve this score
- Raise public artefacts: One published case study would move this the most.
- Raise third party validation: Get a practitioner critique on the record.
- Raise measured outcomes: Attach a number and a method to one shipped result.
Limits
- Inputs are self reported, so the score reflects what you declared, not verified facts.
- This is decision support, not career, legal, financial, psychological, medical or investment advice.
- A high score does not guarantee an outcome. It means the evidence you hold is more consistent.
Why this score? Execution readiness
- Time pressure2 of 5, weight 3Reduce scope per week so the plan survives a busy period.
- Scope spread2 of 5, weight 3Drop two of the three target roles for now.
- Proof debt1 of 5, weight 2Convert learning into one artefact each cycle.
- Feedback loop delay3 of 5, weight 2Ask for critique weekly instead of at the end.
What would improve this score
- Raise proof debt: Convert learning into one artefact each cycle.
- Raise feedback loop delay: Ask for critique weekly instead of at the end.
- Raise time pressure: Reduce scope per week so the plan survives a busy period.
Limits
- Inputs are self reported, so the score reflects what you declared, not verified facts.
- This is decision support, not career, legal, financial, psychological, medical or investment advice.
- A high score does not guarantee an outcome. It means the evidence you hold is more consistent.
Inputs
Current strengths
- Five years of analysis work on data heavy products
- Comfortable writing specifications and acceptance criteria
- Has shipped two internal reporting tools end to end
Missing proofs
- No public artefact that shows AI product judgement
- No documented discovery interview set
- No measurable outcome attached to a shipped AI feature
Constraints
- Around six hours a week outside current work
- Prefers a move without relocation
- Wants to avoid a step back in seniority
Target roles
- AI Product Manager
- Product Manager, data platform
- Technical Product Manager
Next best move
Publish one AI product case study that shows a problem, the option set, the trade off and the measured result.
Top skill gaps
- Model behaviour judgement, evaluation and failure modeshigh
Run a small evaluation on an open dataset and write the failure analysis.
- Pricing and unit economics of AI featuresmedium
Model the cost per request of one feature and publish the sensitivity table.
- Stakeholder narrative for technical audienceslow
Record a five minute walkthrough of one decision and the reasoning behind it.
Proof of work recommendations
- One written case study with the decision, the alternatives and the measured outcome
- One working demo that a hiring manager can open in under a minute
- Six discovery interview notes with a synthesised problem statement
- One benchmark comparing two approaches on the same task
Personal Decision Pack preview
The one page summary you would take into a decision conversation.
Decision context
Move toward AI Product Management
Maya, 29, business analyst moving toward AI Product Management
Goal: Become credible for AI Product Manager roles within six months
- Career Readiness Score
- 55 of 100, Needs more proof
- Evidence Strength Score
- 42 of 100, Needs more proof
- Execution Readiness Score
- 40 of 100, Needs more proof
Evidence available
- Two shipped internal tools, self reported
- Three public job descriptions used to derive the required proof
- No third party validation of the shipped outcomes yet
Unknowns
- How target employers weigh public proof against internal experience
- Whether six hours a week is enough to publish one proof a month
Risks
- Learning content is consumed but no observable proof is produced
- The target role definition stays vague, so effort spreads across three profiles
Next best action
Publish one AI product case study that shows a problem, the option set, the trade off and the measured result.
Seven day sprint
- Day 1Write the target role definition and the proof it demandsOutput: One page role brief
- Day 2Pick one problem worth a public case studyOutput: Problem statement
- Day 3Run two discovery conversationsOutput: Two interview notes
- Day 4Build the smallest demo that shows judgementOutput: Working demo
- Day 5Measure one outcome and record the methodOutput: Result table
- Day 6Write the case study and the trade off sectionOutput: Published case study
- Day 7Ask two practitioners for a critiqueOutput: Two critiques logged
Decision support only. ImpactOS Personal structures your own inputs into an explainable decision view. It is not career, legal, financial, psychological, medical or investment advice, it does not predict outcomes, and every important decision stays yours.
Do not just claim ambition. Build proofs.The proof types that make a next move credible.
Public case study
Medium effortShows judgement, trade offs and a measured result in one artefact.
- What it proves:
- That you can frame a problem, choose between options and report a measured result.
- Best for:
- Career and career switch paths
- First step:
- Write one page: context, options, decision, outcome, what you would change.
Portfolio page
Low effortGives a reviewer one place to verify the claim quickly.
- What it proves:
- That your claims are consistent and easy to check in one place.
- Best for:
- Personal branding
- First step:
- Publish a single page listing three artefacts with one sentence each.
Demo project
Medium effortTurns a claim into something a stranger can open and use.
- What it proves:
- That you can ship something a stranger can open and use.
- Best for:
- Career switch to AI and side project
- First step:
- Scope one narrow use case that works end to end, nothing else.
Customer interview notes
Low effortEvidence that the problem exists outside your own head.
- What it proves:
- That the problem exists outside your own head.
- Best for:
- Founder and idea paths
- First step:
- Run five short conversations and record the exact words used to describe the pain.
Benchmark analysis
Medium effortShows you can compare options on the same measurable basis.
- What it proves:
- That you can compare options on the same measurable basis.
- Best for:
- Career and interview preparation
- First step:
- Pick three options, define two measures, publish the comparison table.
Certification evidence
Low effortBaseline signal only. It supports a proof, it does not replace one.
- What it proves:
- A baseline of vocabulary and fundamentals, nothing more.
- Best for:
- Career switch, as a support signal only
- First step:
- Add it next to one artefact that applies what it covers.
Prototype walkthrough
High effortExplains the reasoning behind the build, not only the result.
- What it proves:
- That you can explain the reasoning behind a build, not only the result.
- Best for:
- Interview preparation and founder paths
- First step:
- Record a short walkthrough covering one trade off you made and why.
Decision LensesFixed critical questions, so a decision is stress tested from more than one angle.
CEO lens
Career: If you only get one bet this year, why is this role the highest return use of it?
Founder: What has to be true for this to be worth two years of your life?
Product lens
Career: Which single proof would remove the most doubt for a hiring manager?
Founder: What is the smallest version that still creates the outcome people want?
CTO lens
Career: Can you explain the failure modes of the systems you claim to understand?
Founder: What part of this is genuinely hard, and are you solving that part first?
Recruiter lens
Career: In thirty seconds of scanning, what proves you can already do the job?
Founder: Who would join you, and what would convince them this is real?
Customer lens
Career: Whose problem do you make smaller, and can they describe it themselves?
Founder: What are they doing today instead, and why is that tolerable?
Cohort pilotsIllustrative preview for schools, bootcamps and career programmes.
Cohort pilots
For schools, bootcamps, incubators and freelance communities.
Help a whole cohort make better career and founder decisions with the same evidence model: a structured diagnostic per person, a seven day sprint they can actually track and a Decision Pack a mentor can read in two minutes.
- A shared readiness language across the cohort
- Evidence and gaps instead of confidence and opinion
- Sprint progress a coach can follow without chasing people
- Decision Packs a mentor can review before a session
No cohort is running yet. Pilots are being scoped during the beta.
Coach dashboard
Layout preview only, not connected to any data source.
Completed diagnostics
48
Sprint progress
41%
Packs shared
12
Cohort readiness distribution
- Ready to move18%
- Close, evidence missing34%
- Direction unclear31%
- Not started17%
Top gaps
- Existing proof strength
- Target clarity
- Network leverage
Top risks
- Evidence is self reported and unreviewed
- Target role named too broadly to prepare for
- No time reserved for the seven day sprint
These figures are placeholders that show the intended layout. No cohort data exists and no customer is represented here.
Launch readinessHow this beta is being tested before wider release.
Launch readiness
Roadmap targets, not achieved metrics.What this beta is trying to prove next, in the open. None of these objectives is reached yet, and none of them is presented as a result.
Next 7 days
TargetReach 30 targeted testers.
Measured by diagnostic completion, not by sign ups.
Next 30 days
TargetValidate repeat usage and premium beta intent.
Measured by saved decisions, returning sessions and qualitative feedback.
Next 60 days
TargetValidate Proof Portfolio creation.
Measured by proof briefs copied and packs shared.
Next 90 days
TargetOpen one cohort pilot conversation.
Measured by a pilot intent conversation, not by revenue.
Decision support only. ImpactOS Personal structures your own inputs into an explainable decision view. It is not career, legal, financial, psychological, medical or investment advice, it does not predict outcomes, and every important decision stays yours.
Give feedback on ImpactOS Personal
This is a public beta concept. Tell us whether this would change a real decision for you.
Give feedback or request private beta access
Two minutes. It shapes what gets built next, and nothing you send here is used as demo content.
Deciding for a team, not only for yourself?
ImpactOS Enterprise applies the same evidence and readiness model to complex software changes.