Research Engineer, Large-Scale Search & Data Systems

Mistral; Paris; On-site; EUR 90,000 - 130,000; Full time

Job summary

Mistral is seeking a Research Engineer to join the newly forming Search and Services Platform team. You will build and operate large-scale search infrastructure for models, from self-hosted Vespa clusters to knowledge graphs and databases served through MCP.

You’ll collaborate with researchers across data platform teams to ensure reliability, scalability, and speed at scale. You will work with Python/Go, maintain data pipelines and interfacing tools for model access, and contribute to

Qualifications

• 5+ years in search engineering, data infrastructure or similar SRE/platform roles.
• Proficiency in Python/Go and software engineering best practices.
• Hands-on experience with large-scale search or data systems (Vespa, ES/OpenSearch).
• Experience with site reliability engineering: incident management, on-call rotations.
• Experience with containerization and orchestration (Kubernetes) and infra-as-code.

Responsibilities

• Design, build and operate large-scale search infra and services.
• Maintain diverse data sources: knowledge graphs and scholarly data.
• Expose interfaces for models through queries and MCPs.
• Build data pipelines for ingest, indexing and serving massive datasets.
• Collaborate with researchers and data platform teams; produce documentation.

Skills

Python

Go

SRE

Large-scale search

Kubernetes

Infra-as-code

Data pipelines

Tools

Vespa

Elasticsearch/OpenSearch

ClickHouse

PostgreSQL

Docker

Kubernetes

Job description

Mistral is seeking a Research Engineer to join the newly forming Search and Services Platform team. You will build and operate large-scale search infrastructure for models, from self-hosted Vespa clusters to knowledge graphs and databases served through MCP.

You’ll collaborate with researchers across data platform teams to ensure reliability, scalability, and speed at scale. You will work with Python/Go, maintain data pipelines and interfacing tools for model access, and contribute to