StartDock logo

Case 02 · Service design, product coaching, and responsible AI

StartDock Product and AI Practice

Improving a high-stakes startup service while helping international founders turn uncertain ideas into testable products and stronger evidence.

StartDock Startup Visa ProgramProduct Owner and UX LeadAmsterdamFebruary 2026 to present

30-second case summary

Problem

Founders must understand a complex visa process while proving innovation, feasibility, capability, and progress.

My ownership

Service improvements, product guidance, founder research, prototype-test planning, and AI workflow design.

Current scope

Acquisition, application review, onboarding, program delivery, founder products, and transition planning.

Evidence boundary

This work is ongoing. The page shows current methods and proposed experiments, not invented outcome metrics.

The product view

A visa program is also a service journey.

The founder experience spans marketing, application, selection, onboarding, product development, program participation, and the transition to a self-employed permit. I treat these moments as one connected lifecycle instead of separate administrative tasks.

Decision 01

Connect the lifecycle

Map the full founder journey so website content, application questions, review feedback, program sessions, and progress checks support the same goals.

Evidence: Founder conversations, team input, recurring program work, and review of real applications.

Decision 02

Make requirements usable

Translate formal criteria into clear prompts and actions founders can understand without losing the accuracy required for review.

Evidence: Pitch-deck reviews, step-by-step plans, onboarding questions, and repeated founder feedback.

Decision 03

Build evidence, not polish

Move founders from presentation edits toward assumptions, prototype tests, market signals, and specific decisions.

Evidence: Product mentoring, MVP planning, prototype-testing support, and pitch preparation.

Workstream 01 · Service ownership

Design the service across frontstage and backstage.

A founder sees pages, forms, messages, sessions, and decisions. The team manages reviews, handoffs, records, contracts, and program delivery. Both sides need the same clear state and next action.

  1. 01

    Find and understand

    Clarify eligibility, program value, costs, expectations, and the evidence founders need before applying.

  2. 02

    Apply and review

    Turn a high-stakes application into a clear review process with focused feedback and visible next steps.

  3. 03

    Join and build

    Support onboarding, recurring sessions, product validation, pitch development, and participation during the program year.

  4. 04

    Show progress

    Help founders document milestones, strengthen evidence, and prepare for their transition beyond the Startup Visa year.

Current improvement areas

Founder-facing

Clearer program information, fewer disconnected forms, stronger preparation guidance, and visible next steps after each review.

Team-facing

Consistent review cycles, defined ownership, reusable feedback methods, progress check-ins, and better onboarding and offboarding records.

Workstream 02 · Founder product development

Turn a promising idea into a testable decision.

Founders often arrive asking for design feedback. I help them step back and define what they need to learn, who can answer the question, what behavior to observe, and which product decision the test should support.

01

Assumption

Name the belief creating the most product risk.

02

Research question

Define what the team must learn before deciding.

03

Test

Choose a prototype, task, participant, and neutral prompt.

04

Decision

Agree in advance what evidence would change the product.

Decision 01

Prototype-testing support

I guide founders through test goals, participant criteria, tasks, neutral questions, observation, synthesis, and the decision after research.

Evidence: A reusable prototype-testing worksheet and live mentoring sessions.

Decision 02

MVP and pitch alignment

I help founders connect the product they can test now with the problem, market evidence, roadmap, and execution plan they present.

Evidence: Current work across learning, fintech, AI infrastructure, logistics, and advertising products.

Workstream 03 · AI-assisted practice

Use AI for coverage and speed. Keep judgment human.

My first broad pitch-review assistant produced dense feedback and sometimes flattened the founder’s voice. I used those failures as research. The current direction separates the workflow into smaller, focused tools with clear inputs, outputs, and review points.

Step 01

Source evidence

Step 02

Focused AI pass

Step 03

Claim check

Step 04

Founder discussion

Step 05

Human decision

Clear authorship

What AI supports, and what I own.

This boundary keeps the work useful, specific, and accountable.

AI support

Coverage, comparison, and alternatives

  • Scan long decks and plans for repeated or missing information
  • Group feedback against defined RVO and IND criteria
  • Generate alternative ways to explain a difficult point
  • Check consistency across slides, claims, and supporting documents

My responsibility

Research judgment and product decisions

  • Decide which problem matters most for the founder
  • Judge whether evidence supports the claim
  • Protect the founder’s voice and avoid generic AI language
  • Choose priorities, recommendations, and the final narrative
Proposed experiment

Test readiness before asking for a full application.

This experiment is a proposal for the service. No outcome is claimed.

Experiment I would run next

Hypothesis

A short evidence-readiness check before the full application will help qualified founders complete stronger submissions with less avoidable rework.

Variants

  • A: Send founders directly to the standard application.
  • B: Start with a five-minute readiness check and tailored preparation guidance.

Primary metric

Qualified application completion rate

Guardrails

Drop-off, time to submit, reviewer effort, founder confidence

Relevance to bol

New tools need shared standards and accountable decisions.

Relevance to bol

Turning recurring work into focused, reusable methods for a team.
Defining where AI can assist and where a person owns the decision.
Learning from generic output and revising the workflow in response.
Coaching others to connect product decisions with research evidence.
Using AI to support analysis while preserving human authorship and judgment.
Working hands-on across product, research, content, operations, and stakeholder needs.