Company profile

Luma Tech Stack

Multimodal generative AI models for video, 3D and creative media

Software Development San Francisco Bay Area, CA 51–200 employees Privately Held

Luma builds multimodal generative models for video and 3D, shipped through its Dream Machine product. The stack is frontier-lab infrastructure: PyTorch and JAX training on NVIDIA and AMD GPUs over InfiniBand/RDMA, with vLLM, SGLang and TensorRT-LLM for inference. Hiring now tilts toward sales, design and go-to-market roles in the UK, Germany and Saudi Arabia, signaling a push from research into enterprise and agency revenue across EMEA.

Tech Stack 108 technologies

Core StackPython PyTorch TensorFlow Kubernetes TypeScript Go AWS Terraform Apache Airflow Rust C++ Ironclad JAX Unity Midjourney Runway Stable Diffusion vLLM SGLang Ray NCCL MPI OCI NVIDIA AMD InfiniBand RDMA DCGM ROCm TensorRT-LLM+76 more

AdoptingCursor

What Luma Is BuildingChallenges and projects from their job posts

◆Challenges

Scaling RL systems across thousands of GPUs stably
Integrating generative AI into production workflows
Removing usage adoption blockers
Defending against reward hacking
Entering the entertainment vertical

▲Active Projects

End-to-end workflows for brand, agency, entertainment use cases
UK sales motion and EMEA account-growth playbook
Automated evaluation pipelines for generative models
Turning agencies into flagship customers
Vertical-specific demo library and reusable templates
Foundational pipeline in a new market

Hiring Activity

Minimal35 roles · 3 in 30d

Department

Engineering

Sales

Product

Design

Marketing

Support

Data

Executive

Seniority

Senior

Lead

Mid

Staff

Manager

Principal

Notable leadership hires: Creative Lead, Tech Lead Manager, Delivery Lead

Companies like Luma

100+ Software Development companies running Python and Ironclad

See the full list

About Luma

Luma is a privately held AI company headquartered in the San Francisco Bay Area with 51–200 employees. It develops multimodal models spanning video, 3D and generative media, and delivers them through products such as Dream Machine for creators and teams. Its go-to-market targets brands, agencies and entertainment companies with production workflows and vertical-specific demos. Engineering runs on Python, Go, Rust, C++ and TypeScript across AWS and OCI, orchestrated with Kubernetes, Ray, Terraform and Airflow. Hiring spans the United States, Germany, Saudi Arabia and the United Kingdom.

HeadquartersSan Francisco Bay Area, CA

Company Size51–200 employees

Hiring MarketsUnited States, Germany, Saudi Arabia, United Kingdom

Frequently Asked Questions

What tech stack does Luma AI use?

Luma uses PyTorch, JAX and TensorFlow for modeling, vLLM, SGLang and TensorRT-LLM for inference, Ray and Kubernetes for orchestration, and NVIDIA and AMD GPUs on AWS and OCI with InfiniBand networking.

What countries is Luma AI hiring in?

Luma is hiring in the United States, Germany, Saudi Arabia and the United Kingdom, with open work focused on building a UK sales motion and EMEA account growth.

What is Luma AI working on?

Luma is building end-to-end generative production workflows for brand, agency and entertainment use cases, automated evaluation pipelines for generative models, and reusable templates and demo libraries.

Similar Companies in Software Development

Other companies in the same industry, closest in size

How this profile is built

Luma's technology stack, projects, and hiring signals are inferred from public hiring and company data — career pages, public listings, and company web presence — then clustered and de-duplicated. Figures are estimates that refresh over time. Read our full methodology →

This is not an official vendor or customer list. It is a technology-adoption signal inferred from public data, intended for B2B research.