At Viamagus, every engineer works with AI every day - not as a side experiment, but as the core of how we build. We're looking for a Technical Architect who has shipped real systems, led real teams, and already treats AI tools as first-class instruments of engineering. You'll own architecture across all client engagements, mentor a team of 15–20 engineers, and shape how an AI-native consultancy delivers at scale.
- Direct access to client CTOs - architecture decisions that actually ship, not slide decks
- Work inside an org where Claude Code, Cursor, and LLM-native workflows are the default, not the exception
- Lead technical direction on diverse engagements: cloud modernisation, AI product builds, enterprise integrations
- Shape Viamagus's engineering standards, internal tooling, and AI practices from the ground up
What you'll do
- Design scalable, secure architectures for client engagements and lead technical due diligence on proposals
- Drive production readiness: incident management, observability, release processes
- Mentor backend, mobile, and cloud engineers - through architecture reviews, code reviews, and retrospectives
- Be the technical face to client CTOs: translate business objectives into architecture decisions and escalate risks with proposed mitigations
- Enforce security-first design including threat modelling, data classification, and AI-specific risks (prompt injection, PII leakage)
- Ensure compliance readiness for ISO 27001, SOC 2, and HIPAA where applicable
What you bring
Engineering foundation
- 8+ years of software development, 3+ years in an architect or lead role; degree in CS/Engineering (Master's preferred)
- Built and owned systems from scratch to production - full lifecycle, not slices
- Multiple integration experiences: third-party APIs, enterprise systems (SAP, Salesforce, ERP), messaging buses, legacy modernisation
- Built reusable platforms, SDKs, and internal tooling adopted across teams
- Technology-agnostic: strong in at least one modern backend stack, one frontend framework, and one cloud platform
- AWS or Azure architecture: VPC design, IAM, container orchestration, cost optimisation
- DevOps fluency: Docker, Jenkins/GitHub Actions, Terraform or CDK
- Performance tuning, distributed tracing, APM tools (Datadog, New Relic, or equivalent)
- AppSec fundamentals: OWASP Top 10, VAPT remediation, secrets management; ISO/SOC 2/HIPAA exposure a plus
AI-era fluency (working knowledge of most of these required)
- Daily use of AI coding tools - Claude Code, Cursor, Copilot, or equivalent - and the ability to articulate where they help and where they fall short
- LLM integration patterns: OpenAI, Anthropic, Gemini, or open-source models; streaming, function calling, structured outputs
- RAG fundamentals: vector DBs (pgvector, Pinecone, Qdrant), embeddings, chunking, retrieval tradeoffs
- Agentic systems: tool use, multi-step agents, LangGraph or CrewAI
- Prompt engineering: versioning, structured outputs, guardrails, handling hallucinations
- AI evaluation and cost awareness: measuring quality, latency, and cost of LLM-powered features
- MCP (Model Context Protocol): awareness of what it is and where it fits
Nice to have
- Open-source contributions or published AI tooling
- Real-time sync experience: CRDTs, Realm, Ditto, or offline-first architectures
- Technical writing - blogs, conference talks, or public GitHub work
- Google, AWS, or Azure certifications (a bonus, not a substitute for depth)