Overview

AI Standards is building sovereign AI systems for enterprise environments, with work spanning AI safety, cybersecurity, quantum defence and on-premises deployment. We’re looking for a hands-on AI

Responsibilities

• Analyse business processes and technical requirements to identify where AI can deliver measurable value and where conventional software is more appropriate
• Define use cases with clear inputs, outputs, data requirements, constraints and acceptance criteria
• Build end-to-end workflows that combine language models, specialist models, retrieval, application logic and human review where appropriate
• Develop agent workflows with explicit responsibilities, tool permissions, state management, execution limits and escalation paths
• Integrate models with approved APIs, databases, document stores and enterprise applications using well-defined interfaces
• Build document ingestion and retrieval pipelines, including parsing, chunking, metadata, indexing and permission-aware access
• Evaluate retrieval quality, source grounding and context selection using representative examples and measurable criteria
• Develop structured model interactions with validated outputs, reusable instructions and regression tests
• Implement timeouts, retries, fallbacks, duplicate-action safeguards and recovery mechanisms so failures are visible and manageable
• Compare models and tools against task quality, latency, operating cost, privacy and on-premises compatibility
• Build evaluation datasets covering task completion, answer quality, tool-use correctness and unexpected or adversarial inputs
• Run controlled pilots, gather feedback and distinguish model limitations from workflow, data-quality or usability problems
• Apply tenant isolation, role-based access and appropriate data-handling requirements throughout integrations and workflows
• Test for prompt injection, unintended data disclosure and unsafe tool execution, collaborating with specialists on mitigation
• Work with AI/ML engineers to improve routing, model selection and escalation based on observed performance
• Package successful workflows for repeatable deployment and support clear handovers to engineering and operations
• Maintain architecture diagrams, integration specifications, evaluation results and operating guides so other people can maintain and extend the system
• Explain technical choices, limitations and results clearly to both technical and non-technical stakeholders

Qualifications

• 2–8 years of relevant experience in applied AI, software engineering, automation, data science or technical solution delivery
• Strong mathematical foundations, particularly probability, statistics and linear algebra, with the ability to interpret evaluation metrics and uncertainty
• Hands-on Python programming skills, with practical experience using APIs, SQL, structured data and Git
• Working knowledge of retrieval-augmented generation, embeddings, tool calling, structured outputs and agent orchestration
• Ability to debug across model behaviour, application logic, data pipelines and external services
• Understanding of access control, data governance and secure handling of enterprise information
• Ability to work through incomplete requirements, document decisions and take ownership of delivery
• Relevant education in mathematics, computer science, engineering or a related field, or equivalent demonstrated expertise