AI Full Stack Engineer

Randstad Technologies Recruitment; Greater London; On-site; GBP 76,000 - 101,000; Full time

Job summary

Randstad Technologies Recruitment seeks an AI Full Stack Engineer (AI & ML Engineer) for a 9-week assignment in London, starting Oct 12, 2026. The role is 40 hours per week on-site at Fenchurch Street with daily rates of 291 PAYE or 388 Umbrella.

You will build secure Python APIs and manage multi-agent, stateful workflows with strong testing and deployment discipline. Key responsibilities include LLM management, prompt engineering, memory management, hallucination control, and observability

Qualifications

• Experience with Python and RESTful API development.
• Strong understanding of JSON Schema and data validation.
• Experience with LLMs and prompt engineering.

Responsibilities

• Build multi-agent, stateful and asynchronous workflows.
• Handle prompt engineering, memory management, guardrails and fallback logic.
• Implement tracing and evaluation frameworks for LLM-based systems.
• Develop secure Python APIs with solid testing and deployment practices.

Skills

Python

API development

JSON Schema

Pydantic

Tools

Google ADK

LangChain

LangGraph

AWS OpenSearch

Job description

Job Overview

Role

AI Full Stack Engineer (AI & ML Engineer)

Location

Fenchurch Street, London

Duration

Oct 12, 2026 - Dec 15, 2026 (9 Weeks, 2 Days)

Hours

40 hours/week

Rate

291 PAYE / 388 Umbrella

Core Technical Stack

• Programming & APIs: Python, RESTful API development, JSON Schema, Pydantic.
• AI & LLMs: Gemini, OpenAI, Claude.
• Frameworks & Orchestration: Google ADK, LangChain, LangGraph.
• Data & Search: AWS OpenSearch (Indexing and Vector embeddings - Highly emphasized), Database Architecture (Intermediate).
• Infrastructure: Public Cloud (GCP, AWS, or Azure), containerization, CI/CD pipelines.

Key Responsibilities

• Agentic Systems: Build multi-agent, stateful, and asynchronous workflows.
• LLM Management: Handle prompt engineering, context/memory management, hallucination controls, guardrails, and fallback logic.
• Observability: Implement tracing and evaluation frameworks, specifically utilizing LLM-as-a-Judge patterns.
• Backend Engineering: Develop secure Python APIs and maintain strong testing and deployment disciplines.