Enterprise Agentic AI Architect
Architect production-grade Agentic AI platforms — agents, LLMs, enterprise APIs, and human workflows — with the rigour of large-scale software engineering.
Role overview
We are looking for an Enterprise Agentic AI Architect who has spent the bulk of their career building large-scale, production-grade software — and has since moved decisively into AI. You understand what it takes to engineer real systems: databases, APIs, distributed workloads, deployment pipelines, performance tuning, and resilience. You now apply that same rigour to systems where AI is one component among many, not the whole solution.
What you will do
- ✓Architect end-to-end Agentic AI platforms
- Design systems spanning agents, LLMs, enterprise APIs, data layers, and human workflows
- Define clear boundaries: what is deterministic logic, what is AI-driven reasoning, what is orchestration
- Own the architecture from whiteboard through production — not just the AI piece
- ✓Apply sound software engineering principles to AI systems
- Separation of concerns, modularity, API contracts, fault tolerance, scalability, security, observability
- Design for idempotency, retries, failure recovery, and auditability from day one
- Ensure agents can safely invoke enterprise systems (CRM, ERP, CPQ, data platforms) without side-effect risks
- ✓Make the right technology calls
- Decide when an agent is necessary versus a workflow, a rule engine, or plain deterministic code
- Choose the right combination of LLMs, RAG, function calling, structured outputs, vector stores, and event-driven patterns
- Evaluate frameworks (LangGraph, Semantic Kernel, AutoGen, CrewAI, etc.) on architectural merit, not popularity
- ✓Build for production quality
- Define evaluation approaches: correctness, hallucination risk, grounding, tool-use reliability, regression testing
- Design monitoring, tracing, and behavioural evaluation for probabilistic systems
- Establish guardrails, agent authority boundaries, and human escalation points
- ✓Lead and enable teams
- Drive technical design reviews and guide senior engineers on implementation
- Build reference architectures, reusable components, and engineering standards for AI adoption
- Translate complex architecture into clear communication for engineering and business audiences
What we are looking for
Foundation: large-scale software engineering. You have spent significant time building enterprise-grade systems end-to-end — not just components. You understand the full lifecycle.
- ✓Strong hands-on experience with Java — J2EE, Spring / Spring Boot, enterprise patterns (comparable stacks such as .NET / C# or Node.js are acceptable, but Java depth is strongly preferred)
- ✓REST APIs, microservices, service-oriented architecture, and enterprise integration patterns
- ✓Relational and NoSQL databases — schema design, query optimisation, indexing
- ✓Distributed systems: messaging, event-driven architecture, async patterns
- ✓Application security, authentication, authorisation
- ✓Performance engineering, caching, scalability design
- ✓CI/CD, containers, Kubernetes, cloud-native deployment (AWS / Azure / GCP)
- ✓Production monitoring, observability, incident response
Progression: AI and machine learning. You have meaningfully crossed into AI/ML — enough to understand how these systems differ from deterministic software and what it takes to operationalise them.
- ✓ML solution architecture: data pipelines, feature preparation, model serving, inference
- ✓MLOps: versioning, deployment, monitoring, retraining
- ✓Evaluating and productionising ML workloads at enterprise scale