Skip to content
Gustavo Ravagnani

01Porto2026

GDO Acceleration Plan

AI ASSISTED

An open sales question turned into an operable opportunity model

GDO Acceleration Plan, project cover

From

How can GDO help brokers sell more?

To

Which opportunities exist, how do we compare them, and what decides the order?

Role
Lead Product Designer
Scope
Product model · prioritization · UX architecture · functional prototype
Team
Coordination as sponsor; a peer Product Designer supported the initiative
Period
Late May 2026 · present

Current state

Work in progress. No prioritization decision has been made through it yet.

Context, problem and intervention

Context

GDO is Porto's opportunity-management product for insurance brokers. Coordination arrived with a business question rather than a feature request: how could the product help brokers sell more? The work started by examining the opportunity structure behind that question, before deciding what GDO should build.

Problem

There was no scoped problem and no shared inventory of what the product could plausibly do about the question. Nothing could be prioritized while the opportunity space itself was undefined.

Intervention

I built the opportunity space as a structured object: a taxonomy linking frictions, proposals, opportunities and initiatives, a prioritization logic, and a functional React application that makes the model operable.

Question, model, prioritization, tool

Three decisions, with the evidence behind each and what it produced.

Decision 01

Treat an open question as a portfolio problem

The request could have been answered with a feature list or a dashboard mockup.

Opportunity mapping diagram: journey stage, observed frictions, opportunity areas and initiative directions.
Opportunity mapping. From scattered journey signals to a structured view of frictions, opportunities and initiatives.
Evidence
Nothing could be prioritized while the opportunity space itself remained undefined and unshared.
Decision
Run the strategic analysis first: Product Value Chain work, an agnostic sales-journey analysis, product-correlation mapping, process analysis and friction mapping, then classify the results into a portfolio.
Trade-off
Weeks passed before anything visible existed for stakeholders to react to.
Consequence
The team gained a structured object to argue about rather than a list of opinions.
Decision 02

Turn the opportunity model into an explicit decision system

A taxonomy could organize the portfolio, but it would not tell the team how to compare competing initiatives.

Decision model diagram: inputs, opportunity structure, evaluation lenses, prioritization signals, decision output and follow-through.
Decision model. How opportunity structure, evaluation lenses and prioritization signals connect into explicit decision states.
Evidence
Applying Kano, RICE and impact-versus-complexity across the initiative set exposed where evidence was sufficient, incomplete or conflicting.
Decision
Structure inputs into KRs, opportunities and initiatives; evaluate them through Design, Business and Technology; then combine Kano, RICE and impact × complexity into explicit recommendation and decision states.
Trade-off
The model reduces discretionary flexibility: incomplete or conflicting inputs remain visible instead of being resolved by intuition.
Consequence
The resulting decision logic is auditable, recommendations can be traced back to the structure, evaluations and signals that produced them.
Decision 03

Ship a working application instead of a diagram

A prioritization model presented as slides is evaluated as an opinion.

Three application screens: overview, prioritization board and initiative detail.
Working application. From a decision framework to an operable tool for comparing initiatives, reviewing evidence and supporting prioritization.
Evidence
The decision states, filters and comparisons only demonstrate their value when someone can operate them.
Decision
Use AI-assisted development to build a functional React, TypeScript and Vite application carrying the real initiative dataset, with technical fields left for Technology to complete.
Trade-off
It is an internal decision prototype, there is no collaboration layer and no production hardening.
Consequence
The initiative was approved for presentation to Business and Design as a working tool.

AI-assisted development workflow

The product model came first. Once the strategy, prioritization logic, information architecture and expected behavior were defined, I used AI to turn that model into working software.

  1. 01 · Define

    • Product strategy & opportunity model
    • Prioritization logic
    • Information architecture & behavior

  2. 02 · Structure

    • Project context
    • Requirements & constraints
    • Acceptance criteria

  3. 03 · Gemini

    • Translate requirements into functional HTML
    • Generate the first executable version

  4. 04 · Figma Make

    • Evolve the HTML foundation into the dashboard
    • Refine interface, hierarchy and interactions

  5. 05 · Review

    • Compare output against the product model
    • Correct inconsistencies
    • Iterate on behavior and visuals

  6. 06 · Output

    • Functional decision application
    • React · TypeScript · Vite

I authored the product model and directed every iteration. Gemini and Figma Make accelerated implementation; I remained responsible for whether the resulting software represented the intended product behavior.

Evidence

Functional React dashboard: portfolio overview, prioritization board and initiative detail.
Functional React dashboard. Application screens used to compare initiatives, inspect signals and support prioritization decisions.
Decision output: prioritization board beside an initiative detail with recommendation, decision status and next steps.
Decision output. An example of how the tool structures a recommendation, review state and next steps.

Interactive prototype

The functional React application running live, carrying the real initiative dataset.

Interactive prototype

This interactive demo is password protected. Enter the access password to open it here.

Need access? Contact me to request access to the interactive demo. Contact me →

Constraints and trade-offs

  • AI accelerated implementation; it did not produce the strategy, the taxonomy or the prioritization logic.
  • I own the strategic approach, decision model, taxonomy, application architecture and dashboard behavior; a peer Product Designer supported the initiative.
  • This front is separate from the lead-distribution engine squad, where I hold design responsibility on a different initiative.

Outcome

The model, the prioritization logic and the application exist and hold the real initiative dataset. What remains is a technology evaluation once the technical fields are completed, and the first prioritization decisions actually taken through it.

The case will carry a stronger claim then. For now it stands on the model and the working tool.

Delivered

  • Complete opportunity model, taxonomy and prioritization logic.
  • Functional React application carrying the real initiative dataset.
  • An AI-assisted build workflow that produced a working tool rather than a proposal.

Not measured

  • Organizational adoption: the tool is not yet in use by Business, Technology or Design.
  • Any backlog or prioritization decision produced through the application.
  • Productivity or sales impact of any kind.

Next case

Chance of Success
All projects