Latest AI JobsBrowse Jobs
Mistral AI logo

Research Engineer, Inference Foundation

Mistral AIMunichScience
Apply on company site
Salary
Not listed
Job type
Full-time
Work mode
Remote
Posted
6 Oct 2026
Status
Open · checked 7d ago

About this role

Research Engineer, Inference Foundation is a full-time position based in Paris at Mistral AI focusing on core inference infrastructure and scaling. This role suits professionals with a strong background in large-scale machine learning services, inference engines, and distributed systems.

Role Overview

The Research Engineer, Inference Foundation role centers on managing and optimizing the core inference engine and its orchestration at Mistral AI. The position involves maintaining production-grade serving stacks, enhancing capacity elasticity, and powering reinforcement learning and post-training loops for frontier models.

Key Responsibilities

Build, fix, and optimize the core inference engine and orchestrator to maximize throughput and minimize latency.
Manage the release process for the serving stack through automated performance gates and progressive rollouts.
Improve serving efficiency by reducing pod startup times, handling cold-cache issues, and implementing smarter caching.
Design and maintain serving topologies that overlap communication with computation and optimize routing.
Construct serving infrastructure to support reinforcement learning and post-training for frontier models.

Requirements

Background in building and operating machine learning or large language model services at scale under latency and availability constraints.
Practical knowledge of inference engines such as vLLM, SGLang, TensorRT-LLM, or similar technologies.
Understanding of inference mechanics including prefill, decode, KV-cache management, batching, scheduling, and parallelism.
Familiarity with distributed and disaggregated serving setups.
Ability to troubleshoot across the stack covering CUDA, NCCL, containers, networking, and storage.
Proficiency in Python for systems tooling alongside PyTorch.
Experience operating infrastructure using Kubernetes.
Knowledge of GPU and networking basics including CUDA runtimes and InfiniBand.

Nice to Have

Proven expertise with vLLM or SGLang through upstream contributions or demanding production usage.
Background in hardware-aware optimization for various model architectures.
Experience serving Mixture of Experts models at scale involving load balancing and expert parallelism.
Skills in CUDA or Triton kernel development and profiling using Nsight.
Production experience using Rust or C++.

Skills and Tools

Python
PyTorch
Kubernetes
vLLM
SGLang
TensorRT-LLM
CUDA
NCCL
InfiniBand
Rust
C++
Triton

Salary and Benefits

Mistral AI provides a comprehensive benefits package that varies by location. Offerings can include healthcare coverage, parental leave, retirement plans, relocation support, wellness programs, and meal or transportation allowances.

About the Company

Mistral AI creates full-stack artificial intelligence solutions ranging from frontier models and developer tools to custom applications. The company partners with enterprises across sectors like finance, healthcare, defense, and the public sector, operating with a distributed team across Europe, North America, Asia, and the Middle East.

How to Apply

Interested candidates should apply directly on the official company careers page.

Skills and tech stack

PythonPyTorchKubernetesvLLMSGLangTensorRT-LLMCUDANCCLInfiniBandRustC++TritonLLM

Questions candidates ask

Where is this job located?

This position is based in Paris.

What technologies are used in this role?

The role involves working with Python, PyTorch, Kubernetes, vLLM, SGLang, TensorRT-LLM, CUDA, NCCL, and InfiniBand, with optional exposure to Rust, C++, and Triton.

What are the primary focus areas of the Inference Foundation team?

The team focuses on optimizing the inference stack at scale, ensuring elastic and fast capacity scaling, and powering the serving infrastructure for frontier model training and reinforcement learning loops.

Are benefits provided with this position?

Yes, Mistral AI offers a comprehensive benefits package that varies by location, including healthcare, parental leave, retirement plans, and relocation support.

How should I submit my application?

Candidates must apply on the official company careers page.

Ready to apply?

The application is on the official Mistral AI careers page.

Apply on company site