2026 - Director, Agentic System Alignment & Reliability - Permanent

About Huawei:

Huawei is a leading global provider of information and communications technology (ICT) infrastructure and smart devices. We are committed to bringing digital to every person, home and organization for a fully connected, intelligent world.

At Huawei, we have two key drivers of innovation: science and technology, and customer needs. Both commercial value and market demands are driving our innovation and determining how we invest in science and technology. Breakthroughs in technology, in return, stimulate customer needs and allow us to create greater value for customers.

About Us:

Backed by Huawei Cloud, Ireland Huawei Research Centre is at the forefront of the emerging 'Agentic Internet' era. Leveraging Huawei Cloud's unparalleled global infrastructure—including the cutting-edge CloudMatrix384 super nodes and the comprehensive Industry AI Foundry—we provide the ultimate computational foundation for enterprise-level AI agents. As a strategic investment by Huawei Cloud, this lab benefits from seamless integration with a secure, highly resilient, and multi-cloud-ready European ecosystem. We are dedicated to pioneering the next generation of autonomous multi-agent systems, transforming world-class compute power into deterministic, trustworthy, and efficient AI architectures for global industries.

About the Role:

As AI transitions from static conversational models to autonomous, multi-step Agentic Systems, the industry's core bottleneck has shifted from "model capability" to "systematic reliability for agent workload." We are establishing a cutting-edge engineering and research hub in Dublin focused on the architecture, deterministic execution, and alignment of Multi-Agent Systems (MAS).

As the Director of Agentic System Alignment & Reliability, you will not be training foundational LLMs. Instead, you will lead a world-class team of distributed systems engineers and alignment researchers to build the "industrial-grade skeleton" around frontier AI models. Your mission is to ensure that complex, multi-agent workflows execute with 100% architectural availability, deep observability, and strict adherence to enterprise logic and safety boundaries.

You will bridge the gap between cutting-edge AI capabilities developed by our global teams and the stringent, zero-tolerance reliability requirements of enterprise deployments.

Potential Projects:

You and your lab will pioneer architectures that solve the "black-box" nature of autonomous agents. Key initiatives include:

Agent Infrastructure (The Deterministic Engine & Network):

• Agent Gateway Architectures: Build robust, neuro-symbolic gateway layers that serve as the definitive security perimeter. Every inbound and outbound agent tool call is authenticated and formally proven to comply with enterprise policies before execution. This makes validation, verification, observability, and provable guardrails first-class architectural components, exactly as emerging best practices dictate.

• Deterministic Orchestration: Architect Directed Acyclic Graph (DAG) based execution frameworks that lock in task dependencies, ensuring multi-agent workflows are linearly predictable and immune to cascading model hallucinations.

• Full-Dimensional Observability: Build tracing and telemetry suites tailored for agentic workflows, providing a verifiable, step-by-step "chain of thought and action" for every system decision.

SRE Agent (Empowering Reliability through STERA)

• Implementing STERA (System-Theoretic Extreme Reliability-risk Analysis), this agent transforms cloud governance. It models complex control loops to identify systemic vulnerabilities that traditional monitoring misses.

• By mapping feedback failures, it automatically translates high-risk scenarios into actionable Chaos Engineering experiments. This ensures continuous resilience against non-linear, cascading failures in our hyper-scale public cloud infrastructure.

Code Agent (The Automated Migration & Synthesis Fleet):

• Lead the infrastructure design for a multi-agent system dedicated to enterprise-scale code migration, modernization, and correctness assurance.

• Build sandboxes execution environments and rigorous testing and verification gates, integrating increasing-strength formal methods (such as property-based testing, model checking, and deductive verification) with human-in-the-loop checkpoints, to guarantee that the synthesized code is safe, functional, and deeply aligned with the intended business logic.

You might be a good fit if you have:

• Extensive Systems Leadership: 10+ years of experience in software engineering, with at least 5+ years managing and scaling high-performing engineering or applied research teams.

• Architectural Rigor: Deep expertise in distributed systems, fault-tolerant architectures, or high-reliability infrastructure. You know how to build systems that fail gracefully and recover deterministically.

• AI/Agentic Intuition: While you don't need to be an ML training expert, you must have a strong mental model of how frontier LLMs (like GPT-5.4, Claude Opus 4.6, or other reasoning models) operate, their failure modes (e.g., hallucinations, context degradation), and how to constrain them programmatically.

• Observability & Tracing Mastery: Experience building or utilizing complex telemetry, distributed tracing, and observability platforms to debug "black-box" systems.

• High Agency & Ambiguity Tolerance: Proven ability to operate in highly ambiguous, zero-to-one environments. You excel at turning abstract research problems (e.g., "make this agent not do bad things") into concrete, testable engineering architectures.

• Prior experience building orchestration frameworks for LLM agents (e.g., heavily modified LangChain, AutoGen, or proprietary MAS orchestrators).

Location: Dublin, Ireland

DUE TO THE HIGH VOLUME OF REPLIES, ONLY CANDIDATES WHO ARE SHORTLISTED FOR INTERVIEWS WILL BE CONTACTED.

Privacy Statement

Please read and understand our West European Recruitment Privacy Notice before submitting your personal data to Huawei so that you fully understand how we process and manage your personal data received.

http://career.huawei.com/reccampportal/portal/hrd/weu_rec_all.html