Interview Gilles Monteil: 27 Years of Animation at Ubisoft

From Mime to Generative AI

Gilles possesses a rare experience in the video game industry: a witness to its complete evolution.

From the era of Ubisoft's first 3D engine to today's mega-productions. From the first makeshift mocap systems to today's generative AIs. From building an animation studio from scratch to franchises that marked a generation.

But what makes his journey truly unique is his dual expertise: trained in mime before touching any 3D software, he spent three decades merging these two worlds.

This interview is longer than my usual format. Each section brings something you won't find anywhere else.

Take your time. But read until the end.

Vanessa - AniMotion

Timeline

1989 → Graduated from Marcel Marceau School, 8 years of theater company 1997 → Hired at Ubisoft Montreuil (Tonic Trouble), 3 months later: Art Director 1998-2001 → Co-created Ubisoft Montreal animation studio with Jean Marc Geffroy

2001-2005 → Prince of Persia, Rainbow Six, Splinter Cell (golden age of sharing) 2005-2007 → "Frankenstein" research team + created Ubisoft's 1st mocap studio 2007 → Assassin's Creed (Altaïr's walk, climbing system)

2010 → Splinter Cell Conviction (animation direction)

2011-2017 → Mentoring animation directors (all Ubisoft studios)

2017-2023 → Realization Director Beyond Good & Evil 2

2023-2025 → Neo NPC project (generative AI for intelligent NPCs)

Questions

The Foundations (1997-2005)

How did your mime background help you become a 3D animator?

I graduated from the "International School of Mime Marcel Marceau in Paris."

Mime as an Art is ultimately quite unknown to the public. Yet, as far as I'm concerned, it's very close to animation.

I finished my 3 years of school in 1989 and tried, with my modest means, to make mime exist as an Art in its own right, trying to move beyond this vision of a one-man illusion show trapped in its own cage.

My master, Etienne Decroux, said: "It's extraordinary that an actor who appears before thousands of people has worked on their voice, diction... but not their body! It's a real disability."

Contrary to the image most people have of mime, I saw it more as a stylization of the body in movement, similar to the stylization of text in a narrative work.

That's exactly what we do in animation.

In 1997, times were very difficult for artists, particularly with a government that wanted to eliminate the intermittent worker status.

At that time, I had heard about motion capture and had a very definite opinion on the subject. For me, this "realistic mime" I was trying to develop was THE answer for mocap actors.

That's when I received this interview posted by Ubisoft looking for dancers, puppeteers, mimes... They wanted to develop their first 3D adventure game, Tonic Trouble, for which they were creating an engine.

3D animation immediately fascinated me! I saw it as a way to direct virtual actors who, for once, did exactly what I asked them to do, without asking a thousand questions!

In my early days, I didn't count my hours because there was so much to learn, and I quickly discovered that I had a technical mind.

I think it was simply because the constraints were so great that without technical tricks, it was impossible to obtain a good interactive experience and animation became an empty shell.

For me, games were similar to theatrical improvisation with this nuance: in improv, the primary quality is listening.

In a game, unfortunately, only the player had to listen to the game and rarely the other way around.

We were so far from everything that was possible with actors, but it was very promising, and it seemed to me that I should try to bring my knowledge of acting and physical performance direction to this field where characters moved on their feet, a bit like in theater.

Tonic Trouble

Why mime helps animation :

Ideally, physical actors and animators have the same goal, but their tools and rules to achieve it are very different.

Actors start with a body that's constantly moving, full of tics and flaws that sometimes hinder the clarity of intention. They must work through repetition with rules of control and physical restraint to achieve the desired performance in a single take.

Animators, on the other hand, start with a static image, and all the rules they follow are motivated by the need to "bring to life," to make things move... They can linger on details to make the composition as perfect as possible.

But ultimately, both teachings are complementary, and I think physical acting allows animators to have a different perspective on the performance to achieve.

When you think about it, animation and cinema were born at the same time. The first animated films copied the slapsticks of the era, which was a golden age for physical actors.

Later, silent cinema was replaced by talkies, and speaking actors replaced silent actors, but animators weren't replaced. In my opinion, they succeeded over time in creating and bringing to life complete actors.

💡 IN PRACTICE:

• Watch slapsticks: Charlie Chaplin "Modern Times", Buster Keaton "The General"

• Study how they use silhouette to tell stories without words

• Apply this principle: your pose must be readable in pure silhouette

• Exercise: Animate an action without props, just the body (opening an invisible door)

Charlie Chaplin

How did you build Ubisoft Montreal's animation studio from scratch?

When I joined Ubi, Didier Poli was appointed Animation Director.

3 months later, the team grew and Didier left Ubi to join Glen Keane on Tarzan.

Dominique Bordenave, who was (besides being an incredible sculptor) the director of the graphics studio at Ubisoft in Montreuil, asked him who on the team could replace him, and he said "Gilles."

From then on, I became Art and Technical Director for animation on the part of the project we were handling (the other part was managed by Michel Ancel's team in Montpellier) and in parallel on the development of A3D.

Ubi had gone public shortly before and was starting to invest and develop in certain countries.

At the time, they had the studio in Romania and the one in Annecy with only programmers, and of course the one in Montpellier.

I participated in setting up the studios in Shanghai and Montreal, where I went to recruit and train the first animators for these studios.

Jean Marc Geffroy, who managed the graphic artists on Tonic Trouble, asked me to join the Montreal studio to take care of the animators in the graphics studio he was directing, which I accepted

The philosophy of sharing

The first 5 years were great. We started from scratch.

Ubisoft under construction had bet on hiring only juniors, which had the effect of creating over time a tight-knit and loyal group that had the merit of facing all problems with enthusiasm.

For my part, I naively pushed animation technology to always push back the limitations of our game characters.

We also had the idea of helping everyone grow by showing everyone's work and sharing feedback.

On the technical side, we factorized tools and development between different projects, which allowed us to move much faster.

During this period, we worked on 3 Playmobil projects, an N64 and PC port of Tonic Trouble, a Batman, a Tarzan, a Jungle Book, a Donald, then Prince of Persia Sands of Time, Rainbow Six and Splinter Cell... with a nice progression curve.

Prince of Persia

Splinter Cell

The Turning Point

It was then decided to put recruitment back in the hands of producers.

Rainbow Six Siege ended up with all the technical animation directors, leaving other projects without technical support.

Personally, I think it was the worst decision made by Ubisoft.

We continued to share because we were still the same people who knew each other, but over time, it became apparent that production strength and vision, as well as follow-up and development, were increasingly lacking.

💡IN PRACTICE:

• Create weekly "animation reviews" with your team

• Share your source files, references, failures

• Document your technical solutions for future projects

• Actively seek feedback, especially critical ones

Technical Innovations (2005-2010)

Can you tell us about the "Frankenstein" team and the creation of Ubisoft's first mocap studio?

Frankenstein : pushing constraints

I later worked with an animation research group called "Frankenstein," where I worked with 5 programmers to try to overcome our production constraints.

Simon Clavet (the programmer who popularized Motion Matching on For Honor) was hired on Frankenstein, and we worked together on a lookat system based on rules derived from mime.

One of my battles was developing an automatic transition system that we implemented on the first Assassin's Creed.

This version didn't work on the main character and was removed because the NPCs (unlike "Altaïr," the main character) no longer slid!

The mocap studio: combining acting and technology

It was also during this period that I set up Ubisoft's first motion capture studio with my friend Claude Comeau.

Then another friend, Jacques Dussault, who needed a bit of a break after a cancelled project, joined us.

We divided the tasks: I handled artistic supervision, Claude handled technical, and Jacques handled administrative.

The first project that used our services and allowed us to cut our teeth was Rainbow Six Vegas.

Motion Capture : Rainbow six Vegas

💡 IN PRACTICE:

• If your studio doesn't have mocap: film yourself as video reference (smartphone is enough)

• Before keyframing, do the action physically to understand the timing

• In mocap: think "physical restraint" - remove tics, don't add them

• Mime technique: look for the "essential form" of the movement, remove the superfluous

Iconic Moments

How was Altaïr's iconic walk in Assassin's Creed created?

On Assassin's Creed, Alex Drouin, who was one of the first animators hired at Ubisoft Montreal, had animated Green Playmobil (Alex on the Farm), Batman, and then became lead animator on PoP Sands of Time.

On AC, he handled all the animations for the main character "Altaïr," and with programmer Richard Dumas, they created the climbing system together.

At the time, we were very impressed by the game "Shadow of the Colossus."

Alex and Richard took up the challenge, even though this development wasn't really in the initial design, but it's what greatly contributed to Assassin's Creed becoming a universally recognized franchise.

I think it's the perfect example of what an animator/programmer duo can bring to a game.

Alex and Sylvain Bernard (who handled the NPCs) asked me to come to the mocap studio to work on a walk for Altaïr.

The reference was the eagle, the stealthy predator...

It was a real pleasure for me to create a custom walk, based on mime: arms like eagle wings, feet like brushes, head-neck contradiction, walk without static accent... all this body vocabulary that allows the birth of a stylized character.

Afterward, Alex retouched the walk because the technique on Assassin's Creed (which he also had on PoP) was to use mocap but re-keyframe the animations to refine their rhythm and silhouette to make them compatible with the game systems and give them back a style between cartoon and realism.

Assassin's Creed : Altair

💡IN PRACTICE:

• Define 3-4 "body words" for your character (e.g., "heavy", "nervous", "noble")

• Apply them to ALL their animations for consistency

• Animal reference: helps find a clear physical signature

• Film yourself "playing" the character, exaggerate traits, then refine

• Anim/prog duo: If you have a feature idea, find a programmer ally BEFORE pitching

Splinter Cell Conviction was a reboot under pressure - how did you handle the animation direction?

I asked Alex to create Sam Fisher's walk for Splinter Cell Conviction, but design changes didn't allow us to keep it.

I had to redo it with another actor so it would fit the design.

Ultimately, 5 different actors were used for Sam Fisher's navigation.

I had to do the jog and "plant and turns" myself because they didn't work once implemented in the game.

After Frankenstein and setting up the mocap studio, I was brought in to take over projects that weren't going well, and that was the case with SC Conviction.

After 2 and a half years, the project was still searching, and I was part of the reboot team.

It was intense but surely the best project I had the chance to work on in terms of focus and organization because we had no room for error and had to make wise decisions without getting lost in the field of possibilities.

For family reasons, I left Montreal in 2011 and moved to Montpellier, but I joined Ubi's editorial team in Montreuil to mentor animation directors from all studios and thus carry a vision and push innovation in animation within Ubisoft studios.

The lesson: What works in mocap or standalone animation doesn't always work once integrated into game systems.

You have to constantly adapt, iterate, sometimes redo everything.

But when you have no room for error, you make better decisions and focus on what's essential.

Splinter Cell Conviction

💡 IN PRACTICE:

• Test in the engine as early as possible, not in Maya/Blender

• Beautiful "standalone" animation ≠ animation that works in-game

• On a reboot/crunch: identify the 3 truly important things

• Be ready to redo: ego has no place facing technical constraints

The Future

You're working on Neo NPC with generative AI - how do you see the future of gameplay animation?

 I joined the Neo NPC project, which aims to use GenAI to create intelligent NPCs.

I know there are many questions about AI and many fears about the implications and changes that will affect our future.

I think that in some areas, we'll need to protect ourselves from impacts, but regarding animation and games, I think it should be a tool that allows us to do things that were impossible before.

These technologies are advancing very quickly because enormous budgets are being invested, and I unfortunately don't think we'll stop progress, but we must be part of it to try to steer it toward ends that will help us be more creative.

For video games, a profitable and fun model remains to be defined, but I can see today a future where games can in turn listen to players' proposals.

Text, sound, images, and video have made and continue to make impressive progress in generation every day.

For 3D animation, it's more complicated because there's much less quality material freely available for learning systems to feed on.

Moreover, writing (prompting) what you want is quite restrictive, and if a picture is worth a thousand words, an animation is worth millions.

I still don't believe that game animators will be replaced, but I think on the contrary that their work could be much more interesting.

Today, a game animator tries to craft a movement with an intention.

They then spend all their time having to adapt their base animation to all the changes that can happen in real-time in the game.

I think AI will be able to handle this adaptation, and animators, instead of making just one intention, will make as many as necessary to create unique characters that will increasingly resemble conscious actors.

The work will surely be different, but the knowledge of how to achieve the desired result will remain the same.

💡IN PRACTICE:

• Learn AI basics now (Motion Matching, ML tools)

• Your expertise in intention and "craft" will be more important, not less

• Focus on creating varied "intention libraries"

• AI will handle technical adaptation, you'll do artistic direction

Thanks to Vanessa for asking me to share all this.

I hope this long text (not AI-generated :D) interested you, and I hope it can generate future debates, why not.

I'll end with this quote from Etienne Decroux: "Everything is permitted in Art, as long as it's done on purpose."

Conclusion

Thanks to Gilles for sharing 27 years of experience with rare honesty.

If you've made it this far, you've just traveled through three decades of gameplay animation in 20 minutes.

What remains, beyond AAA franchises and technical systems, is a philosophy: sharing makes you progress faster than competition.

Animator/programmer duos create mechanics that define franchises.

Technical constraints force creativity rather than stifle it.

And AI, like any new technology, is a tool to create the impossible, not a threat.

Gilles created Altaïr's walk with mime principles over 50 years old.

He's now working on NPCs driven by generative AI.

Between the two: the same obsession with creating "conscious actors."

To learn more about Gilles, here's his LinkedIn:https://www.linkedin.com/in/gillesmonteil/

Now it's your turn!

What part of the interview struck you most?