#EG Engagement Solution Architect at NCS3

About the AI Centre of Excellence You will be part of the AI Centre of Excellence (COE), a centralised platform team that builds, operates, and evolves an enterprise-grade GenAI platform consumed by multiple AI projects across the organisation. The COE operates as a platform product organisation, not a project delivery team. It owns shared capabilities end-to-end — reusable building blocks, common services, agent harnesses, orchestration frameworks, retrieval, governance, and observability. Success is measured by platform adoption across AI projects, reliability, and long-term value, rather than one-off delivery milestones. The role In this role, you are the COE's outbound technical voice. You sit inside the COE but spend most of your time engaging with AI project teams — co-creating GenAI solutions with their architects, helping them adopt the COE's common services and building blocks, and bringing their needs back into the COE roadmap. This is a consultant-architect hybrid role. It is technical, but the primary mode is engagement — helping AI project architects design solutions on top of the platform, running architecture reviews, and driving adoption through credibility rather than authority. You do not own final solution design for any one AI project — that sits with the AI Solution Architect embedded in the project. You also do not build the platform itself — that sits with the Platform AI Architect and Platform AI Engineers. Your unique contribution is the bridge. Key responsibilities • Engage AI project teams to co-create GenAI solutions on top of the COE platform. • Guide AI project architects in applying the COE reference architecture, common services, and building blocks to their specific use cases. • Run architecture workshops, design reviews, and technical deep-dives with AI project teams. • Capture recurring needs across AI projects and feed them into the COE Product Manager and Platform AI Architect for roadmap consideration — "more AI projects need feature X, the COE should build it once and serve it everywhere." • Drive adoption of COE common services across AI projects through evangelism, enablement, and hands-on support. • Act as the trusted technical contact for AI project architects — first call when they hit something they don't know how to design. • Stay current on GenAI applications, agents, RAG, and orchestration patterns so your advice stays relevant.Technical environment Across all roles, you will be working with — or designing for — the following stack. You do not need every item listed; you should be familiar with most and able to learn the rest. • LLM providers: Commercial LLM APIs and self-hosted open-weight models served via vLLM / LLM-d. • Agent frameworks: LangChain, LangGraph, LlamaIndex. • Retrieval: Milvus and Graph DB, with a custom-built ingestion and retrieval framework on top. • Orchestration and harness: Custom-built harness using Skills, eval loops, prompt and context engineering, and multi-agent orchestration on a service mesh. • Runtime: Containerised applications on OpenShift; multi-cloud Kubernetes (AKS, EKS) deployed via GitOps and ArgoCD. • Guardrails: Custom guardrails plus F5 AI Guardrails, Presidio (PII redaction), and OpenTelemetry plugins. • MLOps: Lightweight, focused on LLM fine-tuning and embedding model lifecycle. (No deep / classical ML.) • Observability: OpenTelemetry, Grafana, Prometheus, Loki, Tempo.How we work • Governance by default: Compliance, security, and guardrails are inherited from the platform, not bolted on per project. • Apps lead, infra supports: GenAI is mostly application-layer work with infrastructure underneath. We expect everyone to be apps-strong and infra-literate. • Telco context: We operate in a telco environment. Telco domain experience is a plus across all roles, not mandatory.