Axiomos
Axiomos

Axiomos (Deeptech AI Infra) — Seeking Technical Advisors & Contributor Partners

Adliswil, Switzerland · Hybrid (Office and Remote) · Part-time

About this role

About Axiomos

Axiomos is developing AXIR, a proprietary universal intermediate representation for AI execution. The goal is to decouple AI workloads from vendor-specific hardware and runtimes (CUDA, ROCm, Trainium, TPU, etc.) and enable standardized execution across heterogeneous accelerators.

The project is already incorporated and the MVP core (CPU + OpenCL backend) has been developed independently by the founder. Current work focuses on validation, performance, standardization, and productization.

What we are looking for

We are no longer seeking a co-founder or CTO. The cap table and governance are closed.

We are open to collaborations with experienced technical contributors interested in:

– Runtime & infra engineering (CPU/GPU/OpenCL/CUDA/ROCm/SYCL)

– DevOps / CI / build / tooling

– Compiler/toolchain engineering

– Productization of deep infrastructure

No salary at this stage. Equity is limited, vesting-based, and only considered for high-value technical contribution (non-governance).

Profiles that may fit

– Senior software engineers (systems, infra, runtime, DevOps)

– Compiler / GPU / HPC engineers

– Technical advisors with infra/productization experience

This is a deeptech infra project, not an AI application startup. If you are interested in foundational infra rather than models/applications, feel free to reach out.

Job details

Salary
No salary
Equity
2% - 5%
Experience
Senior (6-10 years)
Work hours
Flexible hours
Education
Master's Degree
Languages
English

Team Roles

Research & DevelopmentSoftware Engineer

Department

Tech

About the Company

Axiomos
Axiomos

AI & Machine Learning · 1 · Neuchâtel, Switzerland

Axiomos’ mission is to build the universal OS for AI, freeing the world from hardware lock-in. Today, over 70% of AI runs on Nvidia GPUs with CUDA, creating dependency and massive costs. With AXIR, our universal intermediate representation, any AI model can run on any hardware. Our goal is to become the invisible yet indispensable infrastructure for global AI, enabling freedom of choice, scalability, and innovation everywhere.