Modular Tech Stack

AI inference platform unifying model deployment across hardware

Software Development Everywhere 51–200 employees Founded 2022 Privately Held

Modular builds an AI developer platform centered on inference optimization and deployment. The stack is deeply systems-oriented—Kubernetes, CUDA, MLIR, LLVM, PyTorch—with heavy investment in kernel-level performance (GPU optimization, SYCL, OpenCL). The company is actively adopting its own Mojo language and tackling fragmented deployment friction: pain points cluster around model serving performance, cold-start latency, and production deployment scaling. Engineering-dominant hiring (18 of 20 roles) with significant senior and lead concentration suggests a team focused on hard systems problems rather than breadth.

Tech Stack 30 technologies

Core StackKubernetes Python C++ Helm Rust Swift Scala AWS GitHub PyTorch TensorFlow Mojo C/C++ CUDA OpenCL Linux Kernel MLIR LLVM SYCL Haskell Clang GCC Dart GCP Azure asyncio NVIDIA Nsight Systems Nsight Compute PyTorch Profiler JAX

AdoptingMojo

What Modular Is Building

◆Challenges

Friction in deploying trained models
Fragmented ai deployment solutions
Rebuilding ai software stack
Deploy ai in production
Customer satisfaction exceeding expectations
Performance-critical ai inference
Fast and scalable inference
Simplifying model deployment
Performance tuning friction
Cold-start optimizations

▲Active Projects

Ai inference kernel optimization
Llm inference platform
Helm charts and kubernetes operators
Cloud inference products
Policy enforcement for ai data usage
High-performance kernel development
Core library development for ai deployment
Building modular's genai platform
Deliver real measurable value of modular platform
Gpu kernel optimization across environments

Hiring Activity

Accelerating20 roles · 20 in 30d

Department

Engineering

Product

Security

Seniority

Senior

Intern

Mid

Lead

Director

Company intelligence

Find more companies like Modular by tech stack, pain points and active projects

Get started free

About Modular

Modular develops an AI inference platform designed to simplify deploying trained models across diverse hardware environments. Founded in 2022, the company operates as a remote-first organization (HQ listed as 'Everywhere') with 51–200 employees based primarily in the United States and United Kingdom. The product targets performance-critical AI workloads where inference latency, cost, and hardware utilization are core drivers. Active projects span LLM inference, kernel optimization, Kubernetes-based orchestration, and cloud inference products, with consistent focus on reducing friction between model development and production deployment.

HeadquartersEverywhere

Company Size51–200 employees

Founded2022

Hiring MarketsUnited States, United Kingdom

Frequently Asked Questions

What tech stack does Modular use?

Modular's stack centers on systems languages and ML frameworks: Python, C++, Rust, CUDA, PyTorch, TensorFlow, MLIR, LLVM, Kubernetes, and cloud platforms (AWS, GCP, Azure). The company is actively adopting Mojo, its own language for systems programming.

Where is Modular headquartered and hiring?

Modular operates as a distributed organization with headquarters listed globally. Active hiring is concentrated in the United States and United Kingdom.