See how candidates think and approach the work this role demands, before the phone screen. We'll build a video challenge from this posting, and you can edit or share it before it goes live.
Captions (Mirage) is building a 3D rendering engine that composites video and graphics layers, a sign that its AI video company is investing in the traditional compositing infrastructure needed to turn generative models into a professional editing and developer platform (Tesseract) rather than just generating clips.
Mirage is an AI video company focused on making creation dramatically easier. Our team tackles some of the hardest creative and technical challenges in generative media.
We get to rethink how people collaborate with AI that can actually do creative work with them—defining entirely new interaction paradigms for AI-native video tools. We make that possible by building multimodal agents that turn creative direction into edits, rendering engines that composite multiple layers of video and graphics, and generative video models trained from the ground up that create new media when needed.
All of this comes together in Captions, our creative workspace for generating, editing, and designing video. The technology behind Captions is now available more broadly: Tesseract gives any AI agent native video-editing capabilities, while developers can integrate our models into their own products through APIs.
Explore our work
Captions — Our flagship creative workspace
Tesseract — Our professional video engine for AI agents
Research — The foundation models we build in-house
Updates — The latest from Mirage
TechCrunch, Forbes, Fast Company — Press
Our Investors
We’re very fortunate to have some the best investors and entrepreneurs backing us, including Index Ventures, Kleiner Perkins, Sequoia Capital, Andreessen Horowitz, General Catalyst, Uncommon Projects, Kevin Systrom, Mike Krieger, Lenny Rachitsky, Antoine Martin, Julie Zhuo, Ben Rubin, Jaren Glover, SVAngel, 20VC, Ludlow Ventures, Chapter One, and more.
Please note that most of our roles will require you to be in-person at our NYC HQ (located in Union Square)
About the role
We want Tesseract to render 3D that is indistinguishable from footage. You'll be the founding engineer for photoreal 3D in Tesseract, our Rust-based, cross-platform graphics and media rendering engine.
Tesseract already renders video, motion graphics, effects and compositing in a color- managed pipeline. You'll build 3D to the same standard: a fast interactive preview for editing, and a physically based final render, path-traced or hybrid ray-traced, that produces film-quality frames for export. Both run from one scene, material and lighting model, so what people see while they edit is what they export.
Scenes are authored by people and by AI agents, so render cost, failure modes and APIs matter as much as image quality. This is a foundational role: you'll make the core architectural decisions and help build the team around them.
What you'll do
Define the architecture for photoreal 3D: how ray tracing, rasterization and denoising divide the work, and how that maps onto Metal, Vulkan, Direct3D and WebGPU.
Build light transport: global illumination, reflections and refraction, soft shadows from area lights, and volumetrics such as haze, fog and light shafts.
Build a physically based model for materials, lights and cameras: energy-conserving materials (metal, glass, clearcoat, sheen, translucency), lights in real-world units, and a camera with exposure, depth of field and motion blur.
Own image quality: sampling, antialiasing, accumulation, denoising and firefly suppression, with frames that stay stable and deterministic across a video export.
Scale scenes to production complexity with acceleration structures, instancing, culling and memory management, for millions of triangles and tens of thousands of objects.
Integrate 3D into Tesseract's video, motion graphics and compositing pipeline, including ACES color management and render passes such as depth, normals, motion vectors and light groups.
Make render cost predictable and rendering robust: performance budgets, profiling, and headless batch rendering on local and cloud GPUs.
Design 3D APIs that work for people and AI agents alike: a clear scene representation, actionable validation errors, fast previews and render cost known before the render.
Work with engine, product and AI teams, and help hire and grow the 3D rendering team.
What makes you a great fit
Deep experience building physically based renderers in production: film or offline renderers, path tracers, or game engines with ray-traced or hybrid lighting.
Strong command of light transport: the rendering equation, Monte Carlo integration, importance sampling, material models and denoising.
Hands-on GPU ray tracing: building and traversing acceleration structures (BVHs), hardware ray-tracing APIs such as Metal ray tracing, Vulkan ray tracing or DXR, and compute-based fallbacks.
A solid understanding of modern rasterization pipelines, and of how hybrid renderers combine rasterization with ray tracing.
Hands-on experience with Metal, Vulkan, Direct3D or WebGPU, including compute shaders, GPU profiling and GPU debugging.
Proficiency in Rust or C++, with strong reasoning about memory, concurrency and performance.
Strong foundations in the mathematics of graphics: linear algebra, geometry, probability and sampling.
An eye for realism: you can look at a frame, say why it doesn't read as real, and fix the cause.
Experience owning complex technical projects and making architectural decisions in ambiguous problem spaces.
Even better if…
You've worked on the internals of a production renderer or engine such as Cycles, Arnold, RenderMan, V-Ray, Karma, Unreal Engine (Lumen or its path tracer) or Unity HDRP.
You've built or integrated ML denoisers, or worked on neural rendering.
You've worked with color management (ACES, OpenColorIO), HDR output or film-look pipelines.
You've built on cross-platform graphics abstractions such as wgpu, including extending them where they lack features such as ray tracing.
You've optimized rendering for Apple silicon or mobile GPUs.
You have experience with Rust, video compositing, or integrating 2D and 3D rendering in creative tools.
Comprehensive medical, dental, and vision plans
401K with employer match
Commuter Benefits
Catered lunch multiple days per week
Dinner stipend every night if you're working late and want a bite!
Grubhub subscription
Health & Wellness Perks
Multiple team offsites per year with team events every month
Generous PTO policy
Captions provides equal employment opportunities to all employees and applicants for employment and prohibits discrimination and harassment of any type without regard to race, color, religion, age, sex, national origin, disability status, genetics, protected veteran status, sexual orientation, gender identity or expression, or any other characteristic protected by federal, state or local laws.
Please note benefits apply to full time employees only.
Answer a McCoy challenge. Companies hiring in engineering see your response and reach out when they want to talk.
About 15 minutes to prepare and 5 minutes to record. Unlimited retakes and a teleprompter help you answer. No AI training, and you can delete your response anytime.
This isn't an application to Captions.
captions:e1da4a65-5b18-4a9b-873e-5b261a350104mirageThis posting hasn't been probed by our closure verifier yet. Stream C runs on a rolling schedule against postings approaching the close-decision threshold.
See how we measure for definitions, or our corrections log for known issues. Found something wrong? Flag a correction.