Yannis Rossi

Yannis Rossi

Technical and gameplay designer. I build VR experiences and systems that mess with how players see things.

Focus
Technical design, gameplay, VR
Background
4 years of programming, BA in Zurich
Now
Master's student, open to roles

Move over the name. The smear is a small cousin of the effect in Block 4.

Selected work · 1 of 3

Block 4

To add: hero clip 10 to 15 seconds. Same shot twice, effect off and then on. No audio needed. This is the most important visual on the whole site.

A horror game set in the ruins of Chernobyl's Block 4. It has no jump scares. It makes the player doubt their own eyes.

I did everything, from level design to research. For the spaces, I studied Soviet-era factories, bunkers and reactors. For the horror, I spent a month in medical journals reading about the visual triggers of hallucinations and the feelings they cause. Then I designed the effect around that.

The core is a real-time datamosh effect. Most games pre-render this kind of glitch, Cyberpunk 2077 and Mouthwashing among them. Mine runs live in the engine, so it can react to what the player does and be reused in other contexts.

Effect pipeline: scene frame, noise, desaturation, character displacement through alpha decay, output frame Scene frame live render Noise break the image Desaturation drain the colour Displacement characters drift, alpha decays Output every frame, in real time
Simplified pipeline. Swap in the exact stages and maths from the thesis.
To add: research boardReference photos of Soviet factories, bunkers or reactors next to your finished spaces.
To add: trigger to feelingA small table from your research. Which visual trigger, which feeling, how the effect shows it.

Selected work · 2 of 3

Build a Pen in VR

Caran d'Ache wanted visitors to understand what's inside a ballpoint pen. I got a list of manufacturing steps, a headset, and a blank page.

I designed the experience from there. I decided what the visitor does at each step, where the machines stand, and how people move around and handle the parts. The pen is scaled up to giant size. You take it apart, move the pieces around, form the spring yourself, and build the full pen again.

The hard part was turning a factory process into something a museum visitor with no VR experience could understand and do with their own hands.

  1. 1Take it apartThe giant pen comes apart piece by piece.
  2. 2Handle the partsPick up and move every component.
  3. 3Form the springShape it yourself.
  4. 4Build the penPut it back together at full scale.

[One decision you made and why. For example, why the machines stand where they do, or something you changed after watching a visitor try it.]

To add: floor planRedraw the layout from memory. Where the machines stand, where the visitor starts, how they move. Your design, drawn by you, not the studio's files.

Selected work · 3 of 3

VR Foosball

I took a rough prototype from an outside company and helped turn it into a product people could actually buy and play.

I redefined the project's goal around playability and replayability. I designed area effects that change how a match plays out. I worked with the dev team on software and latency problems. Along the way this direction became the owner's vision for the product.

What I started with

  • A rough prototype from another company
  • No clear goal for the player
  • Software and latency problems

What it became

  • A defined goal built on playability and replayability
  • Area effects that change each match
  • Latency work done together with the devs
  • Tables in event centres, theme parks and breakrooms
To add: area effectsA simple diagram of 2 or 3 effects. What triggers them and what they change on the table.
To add: latency problemA before and after sketch of the latency issue and how you and the devs solved it.

The builds belong to Innology, so this page shows my design thinking instead of footage.

Tools

Forkline

A branching dialogue editor that reads each NPC line and tells you what answers it needs.

Branching dialogue tends to break in the same few ways. Dead ends. Lines nobody could reach. Questions where every answer was polite, so the player could never push back. Forkline catches those while you write.

Type a yes/no question and it suggests yes, no and a dodge. An open question gets truth, lie, deflect and refuse. A question like "tea or coffee?" splits into the options it names. Every answer carries a tone, and the scene exports to JSON or Ink so it drops straight into an engine.

It also takes a raw script as a PDF or text file. It splits the dialogue into beats that each fit one visual novel text box, finds the choices, builds the branches and can rewrite the lines. Every rewritten line keeps the original next to it, so the writer stays in control.

Try Forkline

Other projects

ASITP

[What ASITP stands for]. A system that scans [which coins or markets] for small price gaps. [How the AI part decides what to flag]. Built in [language].

Keep this about how it works, not about returns.

Room for one moreA game jam, a small prototype, or an editor tool. Leave this out if you have nothing yet.

Writing

Writing

About

Hi, I'm Yannis

I'm a technical and gameplay designer with 4 years of programming behind me.

I did my BA in Zurich. After that I worked at Innology, building VR releases for entertainment centres. I spent almost a year there as product experience owner, on projects like the Caran d'Ache installation and VR foosball.

I'm now doing my master's at [school]. I care most about systems that shape what the player feels from moment to moment, and about horror that works on perception instead of volume.

Email

[email]

LinkedIn

[link]

itch.io

[link]

CV

[PDF link]