Art direction, UI and character animation for an existing medical simulation app

WHAT THIS IS

I am a medical student in Paris. I have built a clinical simulation app used to
train for the OSCE, the practical clinical exam in French medical studies. You
take charge of one patient, from admission to handover, and you are scored the
way an examiner would score you.

To be clear from the start, because the genre name misleads: this is NOT a
hospital management game. No staff, no inventory, no finances, no room building.
One patient at a time. The medical content and the scoring system already exist,
they are mine, and you will never have to write or judge any of it.

WHERE IT STANDS

The product works. It is a web application (Next.js / React / TypeScript) with a
complete game loop, a scoring system, medical content and an automated test
suite. It is not a prototype and it is not an idea on paper.

Visually, it is not there. Sixteen hospital rooms and twenty-four corridors
exist as pre-rendered Blender stills, at decent concept-art level. Patients are
drawn in SVG and recoloured from the case data. Nothing moves — there is no
animation of any kind today.

I have attached five screenshots of the current state, unretouched. Please look
at them before bidding; they say more than this description. The interface is in
French, which is the product's language.

WHAT I WANT

The whole visual package finished, animation included. Not a light restyle — a
product that looks and feels like a real one.

1. ART DIRECTION AND INTERFACE — the main piece, and it is genuinely open.
The visual direction is not decided, and choosing it is part of the job. I am
not handing you a style guide to execute. I need someone who can commit to a
direction and hold it across a dozen screens. One warning: the current accent
colour is orphaned. It was chosen for an earlier illustrated look that no longer
exists, and the game is now made of desaturated grey-blue renders. Treat the
palette as an open question, not something to defend.

2. PRE-RENDERED 3D INTERIORS (BLENDER)
The existing rooms and corridors need materials, lighting and believable
volumes. Important: the 3D is a production tool, not an execution technology. We
render once to a still image; the app only ever loads images.

3. CHARACTERS AND SIMPLE ANIMATION
The drawn patients no longer match the quality of the rooms. I want characters
that do, plus a set of simple animations that make the scene feel alive rather
than frozen.

THE ONE HARD CONSTRAINT — please read this before proposing an approach

Every clinical case needs its own patient. When I write a case I describe a
person: age, sex, build, skin tone, hair. I need to generate that character
myself, in my own authoring tool, without a developer.

And that generated character must inherit every animation already produced for
the game, and fit it perfectly. I am not going to commission an animator each
time I write a case. In practical terms: a shared rig, an animation library
authored once, and new characters that bind to that rig and retarget cleanly.

If your approach cannot deliver that, it will not work for me, however good it
looks. If it can, that is most of what I am buying.

ON THE TECHNICAL APPROACH — genuinely open, and I want your opinion

The product is a web app today. I am not attached to that, and I would rather
have this conversation than shut it down.

I did rule out a real-time engine earlier, and here is the reasoning so you can
argue against it: my 3D is pre-rendered, so a real-time engine would spend a GPU
recomputing images that are already computed, and the web is currently how the
product reaches its users. But that reasoning was made about STATIC decors. Now
that I want animated characters, it may no longer hold — and I know that
skeleton-sharing and animation retargeting are mature, solved features in some
engines. If that changes the calculation, tell me. I would rather be argued out
of a position than defend it out of habit.

So if you believe a different approach is right — Unity, Godot, a web 3D layer,
something else — make the case. I am listening, and a good argument here is
worth more to me than agreement.

Four things any proposal has to account for. These are not preferences; they are
what the project is:

- I BUILD AND MAINTAIN THIS MYSELF, WITH AI ASSISTANCE. That is how the current
application was written, and it is how it will keep evolving between contracts.
It works today because everything in the project is text a model can read and
write. If your approach moves a large part of the work into a visual editor —
scenes, prefabs, inspector wiring, node graphs — say so honestly, because that
directly reduces what I can do on my own afterwards. This is the single point I
care most about after the character constraint, and it may well be the reason
to stay on the web. Convince me either way.
- There is a working application: game loop, scoring, medical content, tests.
I am not saying it can never be rebuilt — I am saying tell me what that costs
and what happens to it, rather than assuming I will fund a rewrite.
- The web is currently how the product reaches its users, alongside a mobile app.
If your approach changes that, say so plainly.
- I author my clinical cases myself, with no developer in the loop. That cannot
become a paid step.

ONE HONEST CAVEAT

Twenty-two buildings and four playable cases today. The structure is built; I am
writing the content case by case, myself. Better said now than discovered later.

BUDGET

I am a student. My budget is limited and I will not pretend otherwise to attract
bids I could not honour.

The scope above is bigger than one contract, so I want to work in lots. The
first would be deliberately small — a proposed art direction applied to one or
two screens — so we can both see how we work together before committing
further. The budget on this listing is for that first lot.

What I can offer besides money: a project that is genuinely finished underneath
rather than another app idea, a clear scope, and someone who knows what he wants
and understands technical constraints. And work whose usefulness is concrete —
medical students use this to prepare for an exam that decides their career.

HOW TO BID

- a portfolio; the work shown matters far more than the CV
- one sentence on what interests you in this project specifically
- what you would do for a first lot, and your rate
- if you think I am approaching this the wrong way technically, tell me why

I reply to everyone.