Lead Architect - Agentic AI

JOB SUMMARY

We’re looking for a hands-on senior engineer who has taken full architectural ownership of enterprise software, someone who designs and builds complete, standalone applications end-to-end, including the database layer, third-party integrations, and enterprise-grade security and governance controls, rather than someone who has mainly worked on top of an existing managed data or ML platform. Genuine, hands-on experience operating a live, production-scale AI system is essential, including responsibility for its reliability, monitoring, incident response, and ongoing improvement, rather than experience limited to early-stage prototyping.

This is a hands-on leadership role: part lead architect, part product owner, part governance lead. You’ll own the product’s architecture from the ground up while also bringing hands-on experience training Small Language Models (SLMs) and directing a developer focused on that work. You’ll be the technical anchor for the product — setting direction, challenging the roadmap where needed, and ensuring the system is secure, reliable, and built to scale for enterprise customers.

JOB RESPONSIBILITIES

Take full ownership of the technical direction of the agentic AI product

Architect and build the product end-to-end, including designing and directly owning the core database layer, rather than relying on a pre-built or managed data platform

Build and own the integration layer connecting to third-party enterprise systems, including authentication protocols, webhooks, rate limiting, inconsistent vendor behaviour, and API versioning

Own the event-driven architecture underpinning the platform’s core workflow and orchestration engine, including queues, workflow orchestration, and idempotency

Implement enterprise-grade controls — role-based access control, single sign-on, audit logging, tamper-evident record-keeping, and self-hosted/on-premises deployment (Docker/Kubernetes), to meet the product’s data sovereignty requirements

Embed application security into every layer of the architecture, from code to deployment

Design the agentic system’s guardrails and safety constraints, manage inference-rate and latency trade-offs, and decide where to use deterministic, rule-based workflows versus AI-driven processes

Take ownership of the operational reliability of the live system — monitoring, incident response, and ongoing improvement, not just feature delivery

Establish and enforce governance operations — data handling, model behaviour, access controls, and change management

Lead a small technical pod, including directing and reviewing the work of a developer focused on training the company’s Small Language Models

Bring genuine business acumen to technical decisions — balancing customer needs, cost, and delivery timelines

Manage requirements and delivery through Jira, maintain the codebase in GitHub, and coordinate UI build-out via Loveable

Produce clear, thorough documentation for architecture, processes, and product decisions

Act as the primary technical point of contact for leadership and the incoming customer base

QUALIFICATIONS

Proven, senior-level experience designing, building, and owning production software end-to-end — this is not an entry- or mid-level role

Track record delivering and operating complete, standalone enterprise applications, rather than building features on top of an existing managed data or ML platform

Deep, hands-on experience designing and directly operating a relational database layer (Postgres)

Proven experience building integrations against third-party enterprise systems — covering authentication protocols, webhooks, rate limiting, inconsistent vendor behaviour, and versioning — this is one of the most important differentiators for this role

Strong background in event-driven systems — queues, workflow orchestration, and idempotency

Hands-on experience implementing RBAC, SSO, audit logging, tamper-evident record-keeping, and self-hosted/on-premises deployment (Docker/Kubernetes) — essential given the product’s data sovereignty requirements

A security-first mindset — able to proactively identify and mitigate application security risks, embedding security into every layer of the architecture

Genuine understanding of the full stack of building an agentic application — guardrails and safety constraints, inference rates and latency trade-offs, and when to use deterministic, rule-based workflows versus AI-driven processes

Experience with Claude Code, Postgres, and Rust (or a strong typed-language background with a demonstrated ability to ramp up quickly across this stack)

Genuine, hands-on experience building and operating a production-grade AI or agentic system — including ownership of reliability, monitoring, incident response, and ongoing improvement, rather than experience limited to early-stage prototyping

Hands-on experience training Small Language Models (SLMs)

Strong working knowledge of enterprise-class systems: scalability, reliability, and security at production scale

Track record of leading or mentoring other engineers

Comfortable challenging specifications and proposing alternative approaches, rather than simply executing requirements as given — able to communicate and debate technical decisions effectively with internal teams and CXOs, not simply agreeing with whatever is asked

Solid project management skills — able to plan, prioritise, and deliver against timelines with minimal oversight

Strong documentation habits and excellent written and verbal communication

Familiarity with GitHub, Jira, and ideally Loveable

Product-minded: comfortable thinking about the end customer, not just the code