AI Architect

Excolo.; Adelaide; On-site; AUD 180,000 - 260,000; Full time

Job summary

Excolo. is seeking a senior AI Architect to own the design, build, and operation of agentic AI systems within production platforms. You will lead architectural decisions, review production code, and raise the bar for engineering excellence.

You'll collaborate with founders, leadership, product, and data science to deliver scalable, secure, and reliable AI capabilities across workflow integrations and multi-tenant environments.

Qualifications

• 10+ years of hands-on software engineering with production ownership of AI systems.
• Strong foundations in distributed systems and API design.
• Experience with secure data handling and multi-tenant environments.

Responsibilities

• Architect end-to-end agentic and generative AI features.
• Define guardrails between model reasoning and deterministic code.
• Design resilient integrations with external platforms and APIs.

Skills

Software engineering

Distributed systems

Python

API design

Security

Mentoring

Tools

LangGraph

MCP

React

Job description

AI Architect Agentic AI Systems | Hands-on Technical Leadership

Adelaide or Melbourne

I’m partnering with an Australian product company that is building AI agents into real production platforms. These agents run inside workflows where getting it wrong has consequences. They’re looking for an AI Architect to own how that capability is designed, built and run.

This is a senior, hands-on role. You’ll set the technical direction, write and review production code, and raise the engineering bar for the team around you. You’ll work closely with founders and leadership, plus product and data science.

What you’ll do

• Architect agentic and generative AI features end to end, including orchestration, tool calling, structured outputs, memory, and state across long-running tasks.
• Decide where model reasoning stops and deterministic code takes over, with strong guardrails around actions that matter.
• Design integrations with external platforms and APIs that hold up under failure, using retries, idempotency, reconciliation and graceful recovery.
• Build evaluation and monitoring in from day one. That means realistic test scenarios, regression suites and production tracing that decide what ships.
• Tune for quality, latency and cost against measurable targets, through model selection, context and caching choices, and execution paths.
• Architect scalable, cloud-native deployments on AWS, GCP or similar.
• Keep things secure, covering multi-tenant isolation, sensitive data, access control, prompt injection and auditability.
• Lead through design reviews, code reviews, clear documentation and mentoring.

What you’ll bring

• Significant software engineering experience (typically 10+ years), including hands-on ownership of generative or agentic AI systems in production.
• Strong foundations in distributed systems, async processing, API design and data storage.
• Deep expertise in Python. Exposure to frameworks such as LangGraph, MCP, or React is highly regarded but not essential.
• Experience building evaluation frameworks for agent behaviour, such as task completion, tool-use accuracy and failure recovery.
• Sound security judgement and comfort working with sensitive data.
• The ability to defend architectural trade-offs and communicate them clearly, in conversation and in writing.

Nice to have

• A background in transactional platforms, enterprise SaaS, regulated or sensitive-data environments, or voice and messaging agents.
• Australian citizenship. Some projects require eligibility for a government security clearance.