Hybrid AI Architect for Enterprise Agent Platforms

Vallum Associates; Greater London; Hybrid; GBP 80,000 - 110,000; Full time

Job summary

Vallum Associates is seeking an AI Architect in London to design agentic AI systems and orchestrate multi-agent architectures. You will define enterprise reference architectures integrating LLMs with core banking and CRM platforms, selecting frameworks and building hands-on with orchestrators.

Applicants should have extensive experience with LLMs, prompt engineering, memory management, and AI quality metrics, plus cloud exposure, preferably Google Cloud Platform.

Qualifications

• Design agentic AI architectures with multi-agent orchestration patterns.
• Define reference architectures for enterprise agent platforms integrating LLMs with core systems.
• Select and standardize frameworks (LangGraph, Google ADK, MCP, A2A patterns).
• Hands-on expertise with agentic frameworks/orchestrators.
• Experience with LLMs, prompt engineering, tool/function calling, memory management.
• Familiarity with AI quality metrics: task success rate, groundedness, containment.

Responsibilities

• Design and implement agentic AI architectures for enterprise use.
• Define reference architectures integrating LLMs with core banking/CRM/risk systems.
• Evaluate and standardize AI frameworks and tooling.
• Hands-on development with orchestrator-based agent frameworks.
• Ensure memory management and prompt/guardrail engineering.

Skills

Agentic architectures

LLM frameworks

Prompt engineering

Google Cloud Platform

Memory management

RAG patterns

Tools

LangGraph

Google ADK

MCP

A2A patterns

Job description

Vallum Associates is seeking an AI Architect in London to design agentic AI systems and orchestrate multi-agent architectures. You will define enterprise reference architectures integrating LLMs with core banking and CRM platforms, selecting frameworks and building hands-on with orchestrators.

Applicants should have extensive experience with LLMs, prompt engineering, memory management, and AI quality metrics, plus cloud exposure, preferably Google Cloud Platform.