Principal Applied AI Architect

About Allexive

Most companies have done an AI workshop or training series. They've got a pilot in flight that nobody's quite sure how to scale. They are AI-aware, and stuck.

Allexive is the AI transformation arm of the Overdose Group. We turn AI-aware organisations into AI-operational ones, working with mid-market and enterprise clients across NZ and Australia by shipping working systems.

The work spans three altitudes:

• Build. Production-grade AI systems shipped into the client's environment. PoC or multi-phase Build Sprint.
• AI Platform Implementation. We stand up a client team's AI operating model from the foundation up: the data layer their agents reason over, the platform in their tenancy, the context architecture their Skills and agents draw from, the starter library of those Skills and agents, and the change management that turns it into how the team works. Converts to monthly retainer.
• Enterprise AI infrastructure. We architect the shared AI plane the rest of a client's strategy runs on: MCP gateway, shared controls, agentic commerce protocols, vendor selection, phased build roadmap to capital committee.

About the Role

The pipeline is outpacing capacity, and the technical surface area is widening across all three altitudes above.

We're looking for a Principal Applied AI Architect to help set the technical direction of the business as we scale into new markets and new services.

This is not a strategy-only role. It is not a pure engineering role either.

Day to day, that means taking a top-line technical seat on client engagements: leading discovery, shaping solution architecture, setting technical patterns, reviewing implementation quality, and guiding the engineers and delivery partners around you.

You will be the person who can turn a messy client problem into a clear technical path. You will know what should be built, what should not be built, what risks need to be called out, and where the team needs to go deeper.

You will also need to stay close enough to the build to be useful. That means getting hands-on when required: prototyping, pressure-testing patterns, unblocking spiky technical problems, reviewing code or architecture, and proving what is possible before a team scales it.

Required Experience

• Applied AI Architecture Background: Designed and delivered real Generative AI systems into mid-market or enterprise environments. Not demos. Not slideware.
• Hands-On Technical Credibility: You do not need to be the main engineer on every build, but you must be able to prototype, review, challenge, debug, and unblock technical work when required.
• Agentic Framework Expertise: Hands-on with Google ADK, Microsoft MAF, Claude Agents SDK, or equivalent agentic development frameworks. Two or more is a strong signal.
• Full-Stack AI Fluency: Tool use, prompt caching, structured outputs, MCP architecture, RAG patterns, context engineering, agent orchestration, and workflow integration.
• LLM Ops Discipline: Evals, observability, prompt versioning, safety patterns, cost and latency optimisation, model selection, and production monitoring.
• Enterprise AI Infrastructure: Understands MCP gateway patterns, agentic protocols, identity, permissions, auditability, security boundaries, and how AI systems operate at scale inside real organisations.
• Core Technical Stack: Strong enough in TypeScript/Node or Python to work directly with engineering teams. Cloud-fluent across AWS, GCP, or Azure, with AWS preferred.
• Solution Architecture: Can translate business problems into technical architecture, delivery phases, risk trade-offs, implementation options, and clear recommendations.
• Professional Services: Delivered across industries with enterprise clients. Commercially aware, delivery-aware, and knows when to flag scope drift before it becomes a problem.
• Communication: Confidently leads senior stakeholder forums, writes with clarity under pressure, and challenges C-Suite constructively without damaging key relationships.

Preferred Experience

• Public technical writing, conference talks, open-source contributions, or visible work in applied AI, agents, automation, enterprise architecture, or AI platform implementation.

Note: If you don't tick every single box but you are an experienced AI Architect, AI Engineer, or technical leader working deeply in applied AI, we'd still love to hear from you.

Key Responsibilities

• Solution Architecture and Delivery Direction (50%): Lead the technical shape of engagements across all three altitudes: Build, AI Platform Implementation, and Enterprise AI Infrastructure. Turn scoped problems into solution architecture, technical patterns, delivery plans, implementation guardrails, and working systems. Direct the engineering capacity around you, review the quality of the build, and stay hands-on with the spiky bits.
• Client Technical Leadership (25%): Own the technical relationship through delivery. Run discovery sessions, working sessions, architecture reviews, alignment gates, and demos. Be the person the client trusts to tell them the truth about what will work, what will not, where the risk is, and what to do next.
• Pre-Sales and Commercial Shaping (25%): Join discovery and scoping calls as Allexive's senior technical voice. Pressure-test feasibility, shape proposals and SOWs with the senior team, identify hidden complexity, and spot engagement-shaped problems inside messy client conversations.

What We Offer

• A senior seat at a fast-moving firm with real clients and real outcomes
• Direct line into the founder and senior team, with no middle layer between you and the work or the strategy
• A front-row view of NZ and AU enterprise AI transformation, across multiple industries
• Salary and equity / profit-share to be agreed based on seniority. Competitive with senior eng + consulting roles in NZ
• Remote-friendly within NZ, with client travel as required
• Real budget for tools, models, and learning. We use what we recommend