1 / 24
Two-person UE4 adventure with Fabien Guillaume (2016–2019): frozen-world exploration, shifting puzzle rooms, Tower of Babel finale. Shipped a vertical slice and teaser; shelved the full game. Never released — but it taught me end-to-end game dev and scope discipline.
Project Babel was the most ambitious thing I have attempted — a two-person 3D adventure in Unreal Engine 4 with Fabien Guillaume, built remotely from 2016 to 2019. We restarted more than once as we levelled up, shipped a playable vertical slice and a teaser trailer, then shelved the full game. It never released. It was still my end-to-end education in game development.
A lone wanderer crosses a frozen world of ruins and silence — no dialogue, no hand-holding. Environmental puzzles block the path: rooms that shift layout, labyrinths that punish assumptions, spaces meant to feel oppressive and beautiful at once. The goal is the Tower of Babel at the heart of it all, where the secret to restoring life on the planet waits at the top.
Fabien and I split everything — concept, blockouts, 3D models, materials, textures, rigging, level design, lighting, and Blueprint gameplay. Remote collaboration worked fine for logic and blueprints; syncing multi-gigabyte art files across time zones did not. That pipeline friction was as much a lesson as the game design.
Not a shipped game — but tangible output: a third-person vertical slice you could run through, a teaser trailer, and a large body of environment art and systems. The slice is where the design was most honest: less about world scale, more about what playing Babel actually felt like.
Unreal Engine 4 and Blueprints for gameplay. Blender and ZBrush for modelling, Substance Painter for textures, SpeedTree for frozen exteriors. Grey-box blockouts before polish — especially for puzzle rooms and labyrinth flow.
The open world was the mistake — not the frozen setting, not the Tower, not the puzzles. Two people cannot sustainably own an open-world pipeline remotely. Every new area multiplied art debt; every restart reset progress; every large asset sync ate time we did not have. We stopped in mid-2019.
If I started again tomorrow, the open world goes first. A tight labyrinth adventure with shifting rooms and one clear goal — reach the top of Babel — would have been the right scope. We proved the atmosphere and the craft; we never proved the product at open-world scale.
This is why an abandoned project sits on my portfolio: Babel was a full-stack game-dev bootcamp I ran on myself. Scope discipline, end-to-end 3D pipeline work, Blueprint prototyping, and the hard rule that remote teams need a pipeline sized to the team — not the dream. Every client project since inherits the same instinct: prove the loop, size the pipeline, then polish.
The teaser trailer is mood and scale — the Tower on the horizon. The embed on this page is the same clip; the carousel shows in-engine work: frozen exteriors, shifting interiors, concept art in Unreal.
Game dev is still in the picture. Project Huldra is the next chapter — solo UE5 forest horror with the scope lessons Babel taught the expensive way.