Senior ML Platform Engineer (Agentic Platform)

We’ve just reached unicorn status with a $150M Series D, accelerating our vision to transform education through human-led, AI-enhanced learning. Today, 100,000+ tutors teach 90+ languages to learners in 180 countries - and we’re only getting started. As a category-defining company, we’re shaping what the future of learning looks like at global scale.

The Agentic Platform team sits within Platform Engineering. We build shared infrastructure and abstractions that help product and engineering teams create and operate reliable AI agents. Our scope includes durable execution, tools, context, memory, permissions, approvals, failure recovery, debugging, and control.

We work closely with GenAI Platform, which provides shared model access, evaluation, observability, and retrieval, and ML Platform, which supports traditional and predictive machine learning. Agentic Platform composes those foundations into a runtime and developer experience for stateful, action-taking AI systems.

We work in small teams with significant ownership, empowering you to directly shape system architecture, technical direction, and platform evolution. Learn more through our Engineering Blog, Tech Radar, and YouTube channel.

What you’ll be doing:

Agentic AI is an increasingly important part of Preply's AI strategy. As a Senior ML Platform Engineer, you'll build the infrastructure that takes agent prototypes to production. You'll solve distributed-systems problems in long-running, non-deterministic workflows that depend on tools, human decisions, and changing context.

This is a hands-on senior role with broad technical influence. You'll write production code, shape platform abstractions and operational standards, and partner with Applied Science, Data, Product, Security and Engineering teams.

In this role, your responsibilities include:

• 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, including MCP where useful.
• Own agent state, context and memory primitives with clear lifecycle, privacy, retention and retrieval semantics.
• Design identity, scoped permissions and human controls for authorization, approvals, escalation and handoffs.
• Develop guardrails for tool access, policy enforcement, argument validation, sandboxing and action limits.
• Provide end-to-end observability, trace replay and scenario-based evaluation across plans, model calls, tool calls, state transitions and outcomes.
• Ship SDKs, templates, local tools, documentation and paved-road patterns so teams can build agents without re-solving runtime infrastructure.
• Evaluate frameworks, workflow engines, and tool protocols while operating the platform with clear SLOs, capacity and cost controls, incident response, and operational standards.

What you need to succeed:

• Proven track record (5+ years) designing, building, and operating backend services or ML infrastructure handling production traffic.
• Strong Python skills and experience building developer-facing APIs/SDKs or shared libraries.
• Experience building GenAI-powered or agentic systems using tool or function calling, structured outputs, and multi-step workflows.
• AI Expertise: You're a forefront user of Agentic AI. You're able to bring hands-on experience and start influencing the team beyond code generation; reserving judgment for architecture, business logic, and security, using AI as a thought partner, and being able to set up quality gates and safety nets to enhance existing engineering workflows and propose AI-native alternatives.
• Experience integrating production services through APIs and events, including authentication, authorization, and secure service-to-service interaction.
• Experience with at least one major public cloud, its services and infrastructure (preferably AWS).
• Hands-on experience with distributed tracing, system telemetry, SLOs, and incident triage.
• Strong communication skills and at least C1 English level, plus willingness to join a paid on-call rotation.

Nice to have:

• Experience with agent frameworks or durable workflow technologies such as LangGraph, OpenAI Agents SDK, or Temporal.
• Experience with agent evaluation, simulation, or production feedback loops.
• Experience designing sandboxed execution environments, delegated authorization for autonomous systems.
• Hands-on experience with MCP Gateway.
• Familiarity with GenAI observability and model gateways, including the boundary between a shared GenAI platform and the Agent Runtime.

Why you’ll love it at Preply:

• An open, collaborative, dynamic, and diverse culture;
• A generous monthly allowance for lessons on Preply.com, Learning & Development budget, and time off for your self-development;
• A competitive financial package with equity, leave allowance and health insurance;
• Not in Barcelona? We offer an attractive relocation package to join us in our Preply Barcelona Hub
• Access to free mental health support platforms;
• Access to Gympass-partnered wellness and gym centers throughout Spain to promote and support well-being and physical health;
• The opportunity to unlock the potential of learners and tutors through language learning and teaching in 175 countries (and counting!).