Senior AI / Machine Learning Engineer

BLACKROC Contingent Pty Ltd; City of Brisbane; On-site; AUD 140,000 - 190,000; Full time

Job summary

BLACKROC Contingent Pty Ltd in Brisbane is seeking a Senior Machine Learning Engineer to advance production ML beyond demos. You will work on a small, highly technical team close to product and ship real AI features that customers actually use.

You’ll explore agentic AI, multi-agent orchestration, retrieval and ranking, MCP, and tooling, turning prototypes into live traffic with strong Python and software engineering depth. Australian working rights are required.

Qualifications

• 5+ years shipping production ML/AI.
• Strong Python programming and deployment experience.
• Experience with agentic AI concepts and multi-agent systems.
• Ability to ship end-to-end from prototype to production.
• Evidence of shipping impact over papers.
• Full Australian working rights.

Responsibilities

• Develop production‑grade agentic and multi‑agent systems.
• Build orchestration frameworks and director/sub‑agents.
• Improve MCP servers, tool‑use and context environments.
• Implement retrieval, vector search, ranking and embeddings.
• Create evaluation harnesses and graders.
• Ensure observability, guardrails and production reliability.

Job description

I'm looking for people who have moved beyond AI demos and are actually shipping things.

The role:

A major global travel business has built a small frontier AI team in Brisbane and is hiring a Senior Machine Learning Engineer into it.

It's deliberately small. Deeply technical. Close to product.

You'll absolutely experiment with new models, frameworks and techniques, but success isn't a paper or an impressive notebook. It's something people actually use.

What are they building?

The team is already shipping production AI and pushing deeper into agentic systems, multi-agent orchestration, MCP, retrieval, ranking and evaluation.

Millions of customers. Live inventory. Huge supplier ecosystems. Complex commercial rules. Real transactions.

What happens when agents can understand a traveller's intent, reason across choices, interact with live systems, coordinate specialist agents and progressively execute more of the journey?

You'll work across:

• Production agentic and multi-agent systems
• Orchestration frameworks, director agents and sub-agents
• MCP servers, tool-use and context environments
• Retrieval, vector search, ranking and embeddings
• Evaluation harnesses, graders and judge models
• Model routing and post-training where it improves the product
• Observability, infrastructure and guardrails for production AI

Most importantly, you'll take prototypes through to real production traffic.

You're not joining a giant AI function to own one microscopic component.

The team is small enough that you'll influence architecture, product decisions and what gets built next. But behind it sits the data, infrastructure, customers and distribution of a major global organisation.

Who gets my attention? Builders.

You'll probably have 5+ years shipping production ML or AI, strong Python and enough software engineering depth to own what you build through deployment and operation.

• You may be deep in Agentic AI, LLMs, Multi-Agent Systems, Retrieval, Ranking, Evaluation, MCP or ML Infrastructure.
• You absolutely don't need every acronym.
• I'd rather see one genuinely difficult system you've built than twelve technologies you've "worked with".
• Strong product instincts matter.
• Engineering judgement matters.
• Evidence that you ship matters most.

You must have full Australian working rights, and your resume is essential.

I unfortunately can't promise everyone who emails me a 'quick chat'. What I can do is personally evaluate what you send.

If this particular role isn't quite right, I'm currently recruiting across around a dozen different kinds of cross-functional agentic AI opportunities in Brisbane, so good people are absolutely worth knowing.