A VR beat'em up set in the Donjon de Naheulbeuk universe, 200 years before the events of the franchise. The player brawls through corrupted taverns in the city of Mliuej, investigating a mysterious threat. My first project as Game Director, with a 25-person team and 5 external partners, over 2.5 years of production.
Three experiential pillars: Believable presence · Immersive physical interaction · Quirky role-playing universe.



I worked in a tight duo with our producer: he owned the backlog and delivery; I owned the creative direction and player-facing intent. Most sprint scope and milestone decisions were made together.
Acting as product owner alongside the producer, I defined the product vision and platform positioning end-to-end: what the game needed to be for its audience, and how to hold that intent across every production, marketing, and release decision.
The project originated internally: a narrative designer had a personal connection with John Lang ("Pen of Chaos"), creator of the franchise. I was given the responsibility of turning that opportunity into a solid concept: designing the pitch, aligning internal teams, and presenting it to the licensor. Once greenlit, I led the CNC pre-production application and a regional Auvergne-Rhône-Alpes funding application.
I designed and presented the external pitch to PlayStation ahead of the PSVR2 release, owned end-to-end: graphic design of the deck itself built around the game's key art, content structured around the IP, the team's track record, and the commercial opportunity, then direct presentation and discussion with the platform team.
I managed the ongoing platform partner relationship with both Sony and Meta: platform requirements, certification coordination, and release positioning across both platforms.
Adapting a beloved French IP for a primarily international VR market, without betraying Naheulbeuk's identity.
Holding the creative thread across the license holder, Artefacts Studio (art & reused assets), Blanktone (sound), Minuit Douze (PR), and two dedicated voice directors.
Identical gameplay on Meta Quest standalone and PC/PSVR2, not just performance parity. No "lite" Quest version.
Reusing and adapting Artefacts Studio's Naheulbeuk visual identity rather than starting from a blank page.
Mid-production, the leads and I identified a budget and scope drift. Rather than absorbing it by reducing quality or stretching the timeline, we made a more radical call: refocus on the VR core, physical interaction and combat.
Cut: meta-progression layer, slot machine, reputation system, procedural NPC encounters in the hub. Kept: procedural enemy spawning and random tavern sequencing, but each level is now hand-crafted around a specific power the player carries.
Initial pillars (Combat, Humor, VR Interaction, Replayability) described what the game contained. As the project matured, I evolved them into pillars describing what the player should feel: Believable presence · Immersive physical interaction · Quirky role-playing universe.
On every debated feature, the question shifted from "do we have everything we promised?" to "does this strengthen presence, interaction, or the Naheulbeuk identity?" It works better in VR, where felt experience matters more than content count.
I pushed the team to invest specifically in the sense of presence, structured around four components:
As Game Director, I set the high-level intent for levels: which player power was worth spotlighting, what a given space needed to convey. The micro level design (pacing, layout, encounter structure) stayed firmly with each level designer, who owned their levels end-to-end. My role was to validate the different aspects as they took shape, through frequent build testing at every stage of creation. That boundary made ownership a constant balancing act, handled through open and continuous communication throughout production.
I also stayed hands-on as a level designer myself on several main and side levels, from paper concept through prototyping and finalization: my role on the levels I touched was always contributor, not decision-maker.
Level design is my first passion in game development, a craft I find my way back to on every project I work on, whenever there's room for it. The condition is always the same: it has to serve the project, never come at its expense.
Each level had to be designed around a single mechanic and put it on display. Two examples I shipped:
The player can no longer grab anything; pillow-flails replace bare hands as the only weapon. Designed as a deliberate difficulty spike: unlike every other level in the game, players can't pick up items to heal mid-combat. That safety net is gone, which forces more careful positioning and makes every hit costly.


The player leaves a trail of poison every time they dash. The mechanic is taught across two levels:



Warmly received on Meta Quest, combat and immersion broadly praised. More mixed on Steam, for two combined reasons: price not adjusted after the rescope, and PC VR players expecting more than what the Quest hardware baseline allowed at that price point.
Once the game was live, I led a structured analysis across three sources: in-game analytics, influencer conversations, and player reviews. Feedback was categorized into ~10 domains (controls, combat, narrative framing, UX, level design, enemy variety, and others). Findings directly fed a major post-launch update and the upcoming PlayStation 5 version.
Community video coverage was a significant data source alongside analytics. Watching players navigate the game unfiltered, with no team member in the room, surfaces blind spots that playtests rarely catch: real friction points, unscripted moments of confusion, genuine reactions to narrative beats.
I also presented the project at the Game Made In France event to engage with potential publishing and platform partners. For me, release is the start of the next iteration loop, not the end of the project.
International-first strategy shaped a premium cast on both languages, with dedicated voice directors: Ryan Highley (Dofus, Road 96, Haven) for English and Gregory Laisné (Re:Zero) for French. Naheulbeuk's humor doesn't translate literally: translators were briefed to rewrite jokes, not translate them, across 9 written languages.
The franchise began as a French audio series, so even within the game's international-first strategy we were set on starting from the French cast. I was personally in the studio for those recordings, conveying the directorial intent to Gregory Laisné and answering the cast's questions on specific lines. An unusual case for a game: the French performance set the tone and rhythm, with English adapting rather than leading.
In the second half of production, I stepped in to support the writing team where the workload required it. As with level design, I wasn't the lead writer: that ownership stayed with our narrative designer, and my role was to lend a hand where it counted.


To celebrate the game's release, I helped organize a dedicated launch concert featuring NaheulBand, Magoyond, and Cécile Corbel on stage. On the day, I was on-site running the play-test setup: letting attendees try the game live, gathering their first-hand reactions, and collecting early feedback in person.
Running live playtests at the concert meant being face to face with players of all ages in a single afternoon: kids discovering the universe for the first time, longtime Naheulbeuk fans, people who had never touched a VR headset. As a game creator, it marks you differently than any analytics dashboard. You read confusion on someone's face before they can articulate it, and you feel the precise moment a mechanic clicks. It was one of the most formative days of the whole production.