Find a story.

Explore Neeed, Vividish, or WGTakt.

All work

02 / Neeed

An interview that builds a service profile.

AI interviews service providers and turns their responses into structured profiles. The next question has to fit the person answering it.

After: the answer changes the next question
In this later test flow, “SaaS products” leads to a question about SaaS clients. The profile preview reflects the captured development focus.

The project

Describe the work. Let the interview follow.

A photographer, a cake maker, and a software engineer should not have to squeeze their work into the same form. Neeed is a service platform built around understanding what someone offers and what another person needs.

The AI interview is one part of a larger system. Answers need to become useful, editable profile information. A model response needs to survive validation and reach the interface. Service matching needs distinctions that a broad category cannot express.

I work across those boundaries: the question people see, the schema that carries the answer, the profile it becomes, and the cases that break the happy path. The September work on adaptive customer forms is still a local iteration in the source record.

01

Start with their words

Freeform service descriptions preserve the person’s own label. Follow up questions should reveal specialisms, constraints, and evidence of their work.

02

Give AI a reliable interface

Normalize responses, validate control shapes, keep understood details editable, and show a retry when the interview cannot continue.

03

Make the profile truthful

An attractive profile must not invent a price, a review, or an availability claim. Structure should help people compare real information.

6 chapters / 9 Apr 2026 onwards

The build, chapter by chapter.

Browse by date
  1. 01
    Profiles and data flow

    A portfolio is part of the answer.

    Different services need different ways to show what makes them worth choosing.

    Implemented
  2. 02
    AI engineering

    The model answered. The app threw it away.

    A response shape mismatch quietly turned a live AI interview into hardcoded questions.

    Implemented
  3. 03
    AI integration

    “Backend” was not a baking service.

    A long software development description exposed both an input limit and a cross language false match.

    Implemented
  4. 04
    AI product integrity

    A price needs a unit. A profile needs evidence.

    A polished generated profile must not turn missing information into a confident claim.

    Implemented
  5. 05
    Adaptive interface design

    Let the answer change the interview. Keep the question still.

    An adaptive form still needs a stable place to read, answer, and understand progress.

    Implemented
  6. 06
    Generative UI

    Let the answer shape the form.

    The next iteration gives the AI a richer interface vocabulary while keeping the person in control of what it understood.

    In progress

The working record

The product, up close.

Original captures, visible details,
and room to look closer.

Another part of the work

WGTakt