Senior ML Platform Engineer — Agentic Runtime

Preply; Greater London; Hybrid; GBP 140,000 - 190,000; Full time

Benefits offered by this job

Lessons allowance

Learning & Development budget

Health insurance

Gym access in London

Job summary

Preply is hiring a Senior ML Platform Engineer for the Agentic Platform in London. You’ll build production-grade infrastructure for AI agents, focusing on durable runtimes, secure tool layers, and scalable memory/state management.

The role collaborates with Applied Science, Data, Product, Security and Engineering teams in a hybrid setup. You’ll lead design decisions, implement runtimes and observability, and shape platform abstractions while owning end-to-end delivery with strong emphasis on

Qualifications

• 5+ years designing, building and operating backend services or ML infra with production traffic.
• Strong Python skills and API/SDK development experience.
• Experience building GenAI-powered or agentic systems with tool calls and multi-step workflows.
• AI expertise and ability to influence architecture, security, and runbooks.
• Experience integrating production services via APIs/events with auth/security.
• Hands-on with at least one major public cloud (AWS preferred).
• Experience with distributed tracing, telemetry, SLOs and incident triage.
• Strong communication and English (C1) level; willingness to join paid on-call rotation.

Responsibilities

• Build a durable agent runtime with checkpoints, retries, timeouts, resumability, and deterministic recovery where possible.
• Create a secure tool layer for discovery, registration, invocation and versioning across internal APIs and external services.
• Own agent state, context and memory primitives with lifecycle, privacy, retention and retrieval semantics.
• Design identity, permissions and human controls for authorization, approvals, escalation and handoffs.
• Develop guardrails for tool access, policy enforcement, validation, sandboxing and action limits.
• Provide end-to-end observability, trace replay and scenario-based evaluation across plans and tool calls.
• Ship SDKs, templates, local tools, docs and paved-road patterns for agents.
• Evaluate frameworks, workflow engines and tool protocols with clear SLOs and cost controls.

Skills

Python

GenAI systems

APIs/SDKs

AWS cloud

Distributed tracing

SLOs & observability

English (C1)

On-call rotation

Communication

Backend design

Tools

OpenAI Agents SDK

LangGraph

Temporal

MCP Gateway

Job description

Preply is hiring a Senior ML Platform Engineer for the Agentic Platform in London. You’ll build production-grade infrastructure for AI agents, focusing on durable runtimes, secure tool layers, and scalable memory/state management.

The role collaborates with Applied Science, Data, Product, Security and Engineering teams in a hybrid setup. You’ll lead design decisions, implement runtimes and observability, and shape platform abstractions while owning end-to-end delivery with strong emphasis on