Principal AI Architect & Engineer

Principal AI Architect & Engineer Location: Brisbane CBD, QLD, Australia Role Type: Full-Time | Senior Specialist Level Domain: Quantitative Intelligence & Agentic Engineering Opportunity Overview: At MERCEL, we operate with a permanent capital base and a multi-decade horizon to steer investment decisions. Our proprietary decisioning platform transforms raw equity market, real estate, sentiment and alternative data into high-conviction decision signals that enable portfolio management, capital allocation and counterparty engagements. We are appointing a Principal AI Architect and Engineer to lead the evolution of our cognitive engine and production data pipelines. This is a hands-on, senior-specialist engineering and architecture lead role. You will own the end-to-end blueprint connecting our data ingestion pipelines - powered by Azure Data Factory, unstructured and Anthropic's frontier AI models - as well as autonomous, Agentic AI systems. You will ensure our core data infrastructure is not a static asset, but a dynamically evolving, self-improving platform backed by enterprise-grade MLOps and DevOps. If you are an elite cloud and AI architect who thrives at the intersection of production systems, frontier LLM orchestration and quantitative automation without corporate bureaucracy, this is your mandate. What You Will Build & Drive: 1. Core Data & AI Architecture Evolution * Continuous Architectural Evolution: Lead the technical design, refactoring, and scaling of MERCEL’s data and AI pipelines. Continuously upgrade core throughput, schema flexibility and evaluation frameworks as new AI modalities emerge. * Hybrid Data Ingestion Pipeline: Architect scalable pipelines using Azure Data Factory (ADF) and custom middleware to ingest structured market metrics, streaming data, and unstructured proprietary notes, documents and research; Design the MCP server environment for future optionality of the AI models we plug into our platforms. * Frontier Model Orchestration: Integrate Anthropic’s Claude model family into core processing loops for high-reasoning tasks, context synthesis, metadata extraction and quantitative indicator generation. 2. Agentic AI Design & Implementation * Multi-Agent Systems: Design and deploy deterministic and non-deterministic autonomous agent workflows (tool-calling, stateful memory, reflective loops, and multi-agent consensus networks) that empower the human in the loop of critical capital allocation and investment decisions. * Reasoning & Execution Engines: Bridge raw data signals with autonomous downstream workflows across asset evaluation, risk modelling and portfolio synthesis. * Evaluation & Guardrails: Implement robust validation layers, hallucination checks, and ground-truth testing protocols to ensure high-stakes quantitative output integrity. 3. DevOps, MLOps & Production Engineering * DevOps Ownership: Drive end-to-end Infrastructure-as-Code, CI/CD automation and deployment pipelines for our proptech and decisioning platforms. * Observability & MLOps: Build deep telemetry and regression testing for non-deterministic LLM pipelines. * Security & Isolation: Maintain strict data sovereignty and zero-trust security boundaries within our private compute environment. Candidate Profile & Skills Required: Experience Level : 5+ years in Enterprise Cloud & Data Architecture; 3+ years dedicated to LLM orchestration, production MLOps or generative AI systems. Qualification Minimum: Bachelor's Degree. Azure Ecosystem : Mid-level proficiency across Azure (ADF) and supporting tools. AWS skills are a nice add-on, but not essential. Pipeline & Models : Hands-on mastery integrating Claude (Anthropic API) and frontier models into custom data transformation and processing pipelines and for large volume context processing. Agentic AI : Proven experience building production-grade agent frameworks (LangGraph, AutoGen, CrewAI, or custom Python-based state engines). DevOps & Infrastructure : Strong fluency in modern CI/CD, Docker/Kubernetes, Infrastructure as Code and automated testing frameworks for AI systems. Mindset : First-principles thinker who balances structural rigor with continuous experimentation. Ownership mindset will get you far. Thrives in high-autonomy environments. Why Join MERCEL? * Direct Capital Impact: Your design and architecture directly drive multi-asset capital allocation and proprietary edge. * Zero Bureaucracy: Direct collaboration with key decision-makers. No steering committees, no corporate fluff - We value high-velocity execution. * Top-Tier Compensation and Autonomy: Exceptional remuneration aligned with specialist mastery and real-world performance, while being allowed the freedom to explore, learn and innovate. * Elite Computing Environment: Access to top-tier compute, modern tooling and unconstrained innovation leeway, working with hands-on specialists. *** Please note that you need full working rights for Australia. We do not sponsor VISAS at this stage.