Senior ML Platform Engineer – Agentic AI Runtime

Preply; Barcelona; Sur place; EUR 120 000 - 170 000; Plein temps

Avantages offerts par ce poste

Relocation package to Barcelona hub

Health insurance

Learning & development budget

Résumé du poste

Preply is seeking a Senior ML Platform Engineer to advance our Agentic AI runtime. You’ll build production-grade infrastructure for durable, stateful agents, tackling long-running distributed workflows with tool calls, context, and memory.

This hands-on role shapes platform abstractions, APIs, and operational standards in collaboration with multiple teams. You’ll own tooling for secure discovery, versioning, and lifecycle management, ensure observability across plans and calls, and drive safety

Qualifications

• 5+ years designing, building, and operating backend services or ML infra in production.
• Strong Python skills and developer-facing API experience.
• Experience with GenAI-powered or agentic systems and tools.

Responsabilités

• Build a durable agent runtime with checkpoints, retries, and deterministic recovery.
• Create secure tool layer for discovery, invocation, and versioning across internal APIs.
• Own agent state, memory primitives, and privacy-conscious lifecycle management.
• Design identity, permissions, and human controls for authorization and handoffs.
• Develop guardrails for tool access, validation, sandboxing, and action limits.
• Provide end-to-end observability, tracing, and scenario-based evaluation.
• Ship SDKs, templates, and docs to empower teams.
• Evaluate frameworks and maintain SLOs, capacity, and incident response.

Connaissances

Backend services design

Python

GenAI platforms

APIs/SDKs

Distributed systems

Cloud AWS

Observability

English (C1)

On-call readiness

Outils

LangGraph

OpenAI Agents SDK

Temporal

MCP Gateway

Description du poste

Preply is seeking a Senior ML Platform Engineer to advance our Agentic AI runtime. You’ll build production-grade infrastructure for durable, stateful agents, tackling long-running distributed workflows with tool calls, context, and memory.

This hands-on role shapes platform abstractions, APIs, and operational standards in collaboration with multiple teams. You’ll own tooling for secure discovery, versioning, and lifecycle management, ensure observability across plans and calls, and drive safety