AI Architect & Developer

My global consultancy client is looking for both an AI Architect & Developer for one of their next generation financial services projects.

AI Developer Position

Contract Length - 6 months (potential for extension)

Location - London (Hybrid 2-days per week onsite)

Rate - £580/day (Inside IR35)

What you'll do

• Build and ship end-to-end AI/ML features, from data ingestion and training to deployment, MLOps workflows, CI/CD, and model versioning
• Develop LLM/GenAI solutions: prompt engineering, RAG pipelines, embeddings, vector search, and inference optimisation (LoRA/PEFT, quantisation, GPU/TPU)
• Own observability across data, models, and prompts; run A/B tests, drive evaluation, and embed Responsible AI practices throughout

What you'll need

• Hands-on LLM/GenAI experience (Gemini or open source) including fine-tuning, RAG pipelines, prompt engineering, and graph-based workflows (ADK, MCP)
• Strong Python, API/microservices development, GCP (Vertex AI, BigQuery), CI/CD, and containerisation (Docker, Kubernetes)

Nice to have: ML frameworks (PyTorch, TensorFlow, Hugging Face), MLOps practices, API Gateway/ISTIO, IAM/data governance, Responsible AI standards

AI Architect Position

Contract Length - 6 months (potential for extension)

Location - London (Hybrid 2-days per week onsite)

Rate - £750/day (Inside IR35)

What you'll do

• Define and own the enterprise AI architecture vision, reference patterns, guidelines, reusable components, and long-term roadmap aligned to business goals, risk posture, and engineering standards across cloud and hybrid environments
• Design secure, scalable AI solutions end-to-end: data ingestion, feature engineering, model training, inference, feedback loops, and MLOps/LLMOps pipelines with CI/CD, versioning, and reproducibility
• Establish integration patterns (APIs, events, microservices) and agentic architectures (multi-agent orchestration, planner-executor, supervisor-worker) to embed AI capabilities into existing platforms and workflows
• Operationalise observability, zero-trust security (BeyondCorp, IAM least-privilege), and model/LLM telemetry — ensuring audit-ready agent interactions, decision provenance, and AI quality metrics
• Collaborate across product, data science, engineering, security, and business stakeholders to translate architecture into high-value solutions, selecting the right frameworks, cloud services, and orchestration tools throughout

What you'll need

• 7+ years designing enterprise AI/agentic architectures using multi-agent orchestration frameworks (LangGraph, Google ADK, MCP, A2A) with hands-on LLM, prompt engineering, and tool/function calling experience
• Deep knowledge of RAG patterns, vector databases, embeddings, API-first integration, event-driven architectures, and GCP (or equivalent hyperscale)

Nice to have: MLOps/AgentOps, model governance, FCA/PRA compliance, regulated financial services, real-time and streaming inference, IAM/VPC security patterns