Workspace sign-in
A possibilityConceptual venture model

An entrepreneurship programme for universities

From a business idea
to a tested venture.

Students turn ideas into tested business concepts through lessons, customer research and practical exercises.

Facilitators teach and review every stage. Students may use their own AI assistant: it helps, but never replaces evidence.

Request a demonstration Read how it works

Any idea is a fair start, from a problem to solve to something new. Students work alone or in teams of two or three, with no business training assumed. Completing the programme does not guarantee a launch.

The story of one venture

Everything begins
with a possibility.

A problem to solve. Something that exists and could be improved. Something new that nobody has made yet. Each is a fair starting point.

The first weeks give the idea enough shape to investigate, without pretending the answers are already known.

Write the claims that could be wrong, and the criterion before the data.

Then the
world answers.

Students leave the classroom to talk with the people they hope to serve. They ask about the last time something happened, not about the idea. They look for workarounds and what those cost.

Every piece of evidence is recorded with its source and the commitment it shows. What contradicts the idea is kept beside what supports it.

Bring back evidence, not reassurance.

A better idea
takes shape.

A claim fails. The problem moves. The offer narrows to what the evidence can carry. A named reviewer reads the submission at the end of the stage and records what was checked.

An evidenced decision to change direction, or to stop, is a full learning outcome.

Keep the reasoning when the direction changes.

Bring the pieces
into the real world.

Who the venture serves and what those people do and spend today. A business model with every number labelled as evidence or assumption. The tests that were run. A launch plan.

The core programme ends with a launch decision the founder can defend, with or without an AI assistant in the room.

Follow one fictional venture through the stages

A classroom system that runs on paper,
with a hosted workspace alongside it.

9stages, each ending in a human review
84lessons: 68 core and 16 advanced
30teaching weeks in two blocks, as planned

01Who it serves

For universities that
teach aspiring founders.

As a full programme, an elective, or one stage inside an existing course.

Students work alone or in teams of two or three on a venture of their own. No prior business training is assumed. The lessons are applied to that venture, tested against the real world and reviewed by a person.

  1. I

    Solve a problem

    A need that people already act on. The method asks what they do and spend about it today, before anything is built.

  2. II

    Improve something that exists

    A service, product or process that falls short. Students investigate the current alternative and the cost of its failures.

  3. III

    Create something new

    A possibility no one has asked for yet. Students ask what made a past occasion valuable and what people actually chose, without leading them to a pain they did not report.

02The method

Evidence
before opinion.

Students learn a method in short lessons, apply it to their venture, investigate real customers, decide against criteria they set beforehand, and act on the decision with a low-cost real-world test.

The result feeds the next lesson. What a student saves stays on their record: each worksheet with its preserved first attempt and every saved revision, the AI use note, each submission as it was reviewed, and the venture’s claims, findings and decisions. A later briefing can read the criteria saved earlier; nothing is copied into a later worksheet automatically.

The lesson. The fieldwork.
The written decision.
One venture record.

What every lesson contains

  • A reading, and a worked example on a fictional case
  • A guided practice with a model answer that opens after the student’s own saved attempt
  • Knowledge checks with explanations
  • A worksheet applied to the student’s own venture
  • AI support with its limits, and an equal route without AI

What students practise

  • Writing claims that could be wrong, with the criterion set before collecting data
  • Interviewing and observing without leading, and recording every item with its source and commitment level
  • Labelling every key number in a business model as evidence or assumption
  • Running small tests with thresholds written in advance
  • Making a written decision, and defending it without AI help

Three records, never one score

Learning
What each student can now do, recorded for each founder separately. A stop reached honestly earns full learning credit.
Venture evidence
The claims, criteria, findings and decisions, with what each item can and cannot show.
Commercial results
Sales, costs and traction, kept apart from the other two so that neither a launch nor a stop decides a grade.

Nine stages. One question each.

Each stage ends in a human review that records what was checked. Stages 7 and 8 continue with ventures that are trading, because their lessons need at least three months of real operating data. They are scheduled separately from the 30-week core calendar.

  1. 0

    Explore an opportunity

    What need, improvement or new possibility is worth my next few weeks, and why me?

  2. 1

    Founder and opportunity

    What am I trying to do, with what means, and what do I believe that could be wrong?

  3. 2

    Customers and value

    Does a defined group really have this need, and what do they do and spend about it today?

  4. 3

    Solution and value

    Which solution creates value that is clearly different, and how will we test it?

  5. 4

    Business model

    Can this become a viable, responsible business on paper, and which assumption could sink it?

  6. 5

    Prototype and tests

    Will real people act: sign up, use, pay?

  7. 6

    Sell, form and launch

    What does it take to sell, operate and launch responsibly? Two parts: go-to-market and sales; formation, AI and launch.

  8. 7

    Traction and operations Advanced

    Can we keep customers month after month?

  9. 8

    Growth readiness Advanced

    What limits sustainable growth, and what can the venture sustain?

Built on established entrepreneurship practice, including customer discovery, business model design and evidence-led experimentation, taught in Naventros’s own wording with its sources credited in each lesson.

03In practice

One idea. Six reviews.
A claim that failed.

CarryWell is a fictional teaching case: two students on a Dubai campus, taken from Stage 0 to Stage 5 in the example journey you can download below.

Layla and Omar, their campus, the people they interview and every figure are invented to show how the method works. Nothing here is market data or a student result.

The stage test asks whether the work was done properly: criteria before data, raw records kept, a decision memo, a record of AI use. Whether a test succeeded is recorded separately. Progress can mean continuing, revising, pausing or stopping.

Why now

Build a business.
Learn to lead with AI.

Practise setting direction, testing ideas, checking AI’s work and making decisions you can defend.

Naventros builds these habits on a real venture, where a weak decision has a visible cost.

  1. I

    Set the goal

    Define success before collecting data.

  2. II

    Delegate what suits

    Hand over tasks, never the conclusion.

  3. III

    Use judgement

    Use the evidence to decide, and explain why.

  4. IV

    Verify the quality

    Check AI suggestions against reliable sources.

Context: Google Cloud, AI agent trends 2026, pages 9 to 14 and 39 to 46, an industry report on business use of AI. It is not an evaluation of Naventros, and Naventros is not affiliated with Google.

04Your AI Partner

The student attempts.
The AI proposes.
The student decides.

Your AI Partner is the assistant a student already uses. The workspace prepares a task briefing; the student copies it into that assistant by hand and pastes the reply back. No AI model runs inside the workspace.

Every exercise has an equal route without AI. Synthetic people and figures are never evidence. Below, one interview guide from the sample lesson passes through the six steps. Scroll, or use the step controls.

Attempt first
1 of 6

An illustration of the workflow with the sample lesson’s fictional case. No AI service is contacted by this page.

Step 1 of 6

Attempt first.

The student reads the lesson, saves the practice and drafts the first interview guide in their own words. The model answer and any AI critique open only after that saved attempt.

The venture
MarketLink, a fictional ordering service for home food sellers, the worked example in the sample lesson.
The first guide
Four questions. Two use forms the lesson bans: “would you” and “how much would you pay”.
The rule
Draft first. The assistant critiques the guide; it never writes it.

Step 2 of 6

A briefing, copied to the student’s own assistant.

Naventros runs no AI of its own. For each exercise the workspace prepares a task briefing from the lesson and the saved attempt. The student reads it, copies the text into the assistant the institution permits, and later pastes the reply back into the workspace. The peer route prepares the same briefing for a peer or facilitator critique.

What goes in
The lesson’s rules, the draft exactly as saved, permitted venture context, and a return format.
What stays out
Fields marked private, the student’s notes and reflections, other students’ work, and any source whose owner did not permit AI analysis.
Where it runs
Outside the workspace. The workspace holds no AI provider key and calls no model.

Step 3 of 6

Proposals return, field by field.

The pasted reply is read against the worksheet: a heading that names a field becomes a proposal for that field, shown beside the student’s own text. Fields the student owns, such as the conclusion, the decision criterion and the reflection, are never filled from a reply; anything said about them is listed under suggestions. Nothing is written into the worksheet until the student decides.

Flagged
“Would you use an ordering app?” is hypothetical. “What features do you want?” collects ideas, not evidence. “How much would you pay?” is a hypothetical price.
Suggested
Ask about yesterday’s orders. Ask what the seller pays anyone or anything today. Screen first, then ask about the most recent occasion.
Also proposed
A new 1-to-10 satisfaction question.

Step 4 of 6

The student accepts or rejects, one by one.

For each proposed field the student chooses Accept, Partly or Reject and writes a reason; nothing is applied until every proposal has a decision. Accepted text goes into the editable draft. The decisions and reasons are written as lines into the AI use note, which is saved with the worksheet.

Kept
Three rewrites and one deletion.
Rejected
The satisfaction rating. A rating is stated opinion; the guide already asks for behaviour.
Still needed
Real interviews with consent recorded. A practice persona is synthetic and never enters the evidence record.

Step 5 of 6

The AI use note.

One free-text note per worksheet, which the student edits: what the assistant did, what was kept, what was rejected and why, what was checked. The workspace adds a line for each decision, and a worksheet cannot be submitted without the note.

A valid note
“No AI used.” The peer route and the AI route are equal.
Who sees it
The facilitator, with the submitted worksheet and the preserved first attempt; the note feeds the AI accountability part of the stage review.
What it is not
Evidence about customers. An AI answer verifies nothing and approves nothing.

Step 6 of 6

A person reviews.

The facilitator opens the submission as it was sent: the worksheet, the AI use note and the preserved first attempt. Feedback follows the schedule the cohort agrees: continue, which opens the next lesson in a guided cohort, or revise with guidance.

Every worksheet
Facilitator feedback, on the cohort’s agreed review schedule.
End of the stage
A named reviewer examines a fixed submission and records the decision and what was checked.
Never
The workspace does not issue a stage decision, and a better question does not establish demand.

Context with boundaries

The briefing can carry the lesson, the saved draft, the earliest recorded decision criterion and permitted findings. Private fields, the student’s notes and other students’ work stay out. The student reviews what is shared.

Suggestions stay suggestions

An AI answer does not verify a finding, replace customer research or approve a stage. The student’s conclusion, decision criterion and reflection are never drafted by the AI.

The same path without AI

Read the lesson, do the research, complete the worksheet and submit. A peer or the facilitator critiques the guide against the banned list. Evidence standards and human feedback are the same.

05Human review

A person decides.
The next lesson opens
when they do.

Feedback on every worksheet comes from the facilitator, on the schedule the cohort agrees. The decision at the end of each stage is made by a named reviewer against a fixed submission.

In a cohort set to guided progression, a student sees the first lesson; each next lesson opens when a facilitator continues the student’s current submission. Feedback marked revise opens nothing. Open lessons never close, and staff are never gated.

Stage 2 · Customers and value · a fictional student in a guided cohort

  1. Continued5R Reveal: desk research with dated sources, value leaks and outliers
  2. SubmittedInterviewing without bias: the Mom Test principles
  3. Opens when your facilitator continues youFinding and recruiting interviewees ethically; consent
The facilitator reads the worksheet and writes:

An illustration. Nothing on this page is saved or sent.

06A lesson, right here

Interviewing
without bias.

Stage 2, Customers and value. The lesson where most founders first meet the difference between being encouraged and learning something true.

Try the rewrite routine, then read the complete participant lesson, its worksheets and the teaching slides without leaving the page. No sign-in. This page does not save or submit anything.

From a question that invites politeness
to one that produces a story.

Each question below invites the person to be polite or to imagine the future. Reveal the question the lesson asks instead, and notice that each one points at a specific past occasion.

Read the complete sample lessonParticipant lesson, worksheets and teaching slides. Model answers are in the facilitator guide and are not shown here.

Loading the lesson…

07For universities

What a university
receives.

A classroom system that runs on paper, so a university can start without new software, and a hosted workspace with the same lessons alongside it.

First deliveries are led by Kalle Sarraj with university staff alongside, so that staff can take over. Before any fieldwork the university completes a short delivery record.

The classroom package

  • 84 lessons with worked examples, guided practice, model answers and knowledge checks
  • Teaching slides for every stage
  • Worksheets and 32 blank forms
  • The Methodology and Fieldbook
  • Facilitator guides with session plans, failure modes and worked review decisions
  • Calendar, delivery record, agreement schedule and certificate template
  • The UAE and Dubai country addendum

The hosted workspace

  • The same 84 lessons; students read, practise, submit worksheets and receive facilitator feedback
  • Guided or open progression, chosen per cohort
  • Students work with or without AI; the workspace prepares briefings and calls no model
  • Invitation only. No public sign-up, no email, no AI model inside
  • Facilitators see the submitted work of their own cohorts; private lesson records stay with the student
  • Students can export their venture record

A university may deliver Naventros entirely on paper, or use the workspace for some cohorts and not others.

Scope and calendar

Thirty programme weeks in two blocks of fifteen, with a review at the end of each stage. University breaks are placed by the institution. A shorter engagement is possible: one stage, or the Stage 2 customer evidence sprint inside an existing course.

08Documents

Read it in
your own time.

The partner folder, as PDF. About ninety minutes to read it all; five minutes for the two-page overview.

The facilitator guide for the sample lesson contains model answers and is shared with teaching staff in a delivery discussion, not here. Optional downloads, no registration.

Read the two-page overview PDF, two pages. The rest of the folder opens below.

09See it for yourself

Request a
demonstration.

A supervised demonstration: the sample lesson taught as in class, a live walk through the workspace with a fictional cohort, and a conversation about the scope that fits your calendar.

info@naventros.com

Already part of a programme? Open the workspace.

Fields marked * are required.

Your details stay in this page until you choose to send them. Sending shares them with Naventros at info@naventros.com, through its email services, to respond about your programme. This form does not create a learner record. Do not include student records or confidential venture information.

You can review the details, then send them through your email app. Nothing is sent automatically.

Before you begin.

Do students need an existing business idea?

They can start with a problem to solve, something that exists and could be improved, or something new to create. Stage 0 asks what is worth their next few weeks, and why them.

Does every student finish with a launched business?

No. The core programme ends with a launch decision the founder can defend. An evidenced decision to change direction or to stop is a full learning outcome, and completing Naventros does not guarantee a launch, revenue or investment.

Is an AI subscription required?

No. Your AI Partner is the assistant a student already uses, and every exercise has an equal route without AI. The university decides which AI services are approved for students, or that none are used.

Is the digital workspace required?

No. The classroom package runs on paper. A university may use the workspace for some cohorts and not others; it is invitation-only and runs no AI model.

Can our staff lead the programme?

That is the intention. First deliveries are led by Kalle Sarraj with university staff alongside. Before independent delivery a facilitator teaches an observed lesson, guides a difficult exercise and assesses shared cases consistently with other facilitators.

Is academic credit or accreditation included?

No accreditation or academic credit is implied. Each institution sets its own grade weighting; Naventros keeps learning, venture evidence and commercial results as three separate records.