About The Role FDM is a global business and technology consultancy seeking a Prompt Engineer to work for our client within the Financial Services sector. This is initially a 6-12 month contract with the potential to extend and will be a hybrid role that will be based in Edinburgh.
Our client is seeking IDP Prompt Engineers to design, develop and optimise AI-driven document-processing solutions within an enterprise AI Hub.
The role combines prompt engineering, Generative AI, OCR and Intelligent Document Processing to improve how structured, semi-structured and unstructured documents are classified, understood, validated and processed.
You will work with technologies including Azure AI Document Intelligence, Azure AI Foundry, Azure OpenAI Service and Tungsten TotalAgility. You will translate business requirements into secure, reusable and scalable document-intelligence capabilities for regulated, document-intensive business processes
Design, develop, test and optimise prompts for document understanding, classification, data extraction and validation.
Configure and optimise Azure AI Document Intelligence models and Tungsten TotalAgility workflows.
Develop hybrid OCR and Large Language Model processing approaches for structured, semi-structured and unstructured documents.
Support complex extraction requirements, including handwritten content, tables and variable document formats.
Define confidence thresholds, validation rules, exception-handling processes and prompt guardrails.
Evaluate extraction accuracy and monitor performance against agreed success measures.
Diagnose extraction failures and continuously improve prompts, models and processing workflows.
Establish reusable prompt libraries, templates, testing standards and quality-assurance processes.
Document design decisions, assumptions, test results and lessons learned.
Support AI governance, security, compliance, threat modelling and model-validation activities.
Contribute to operational readiness, service transition and knowledge transfer into BAU support.
Present technical recommendations and performance findings to business and technical stakeholders.
Provide technical leadership, coaching and guidance on the safe and effective use of Intelligent Document Processing capabilities
About You Required Experience:
Strong experience designing, testing and optimising prompts for LLM-based document-understanding and data-extraction use cases.
Hands-on experience with Azure AI Document Intelligence, Azure AI Foundry and Azure OpenAI Service.
Strong understanding of Intelligent Document Processing, document classification and automated data extraction.
Knowledge of Retrieval-Augmented Generation, grounding techniques, guardrails and model-output evaluation.
Experience establishing prompt-testing and prompt-quality-assurance approaches.
Strong knowledge of OCR, ICR, handwriting recognition, table extraction and confidence scoring.
Experience designing validation rules, exception workflows and human-in-the-loop processes.
Ability to translate business requirements into effective AI interactions, behaviours and technical solutions.
Experience working with JSON, XML, REST APIs, workflow orchestration and prompt-evaluation frameworks.
Understanding of AI governance, security, compliance, risk management, threat modelling and model validation.
Strong analytical and problem-solving skills, with the ability to diagnose model and extraction failures.
Strong stakeholder-management, written communication and presentation skills.
Python development experience.
Experience with Microsoft Power Platform.
Hands-on experience with Tungsten TotalAgility, including classification, extraction, validation and OCR configuration.
Experience delivering enterprise-scale AI-driven document automation.
Previous experience within financial services, banking, insurance, transfer agency or another regulated environment.
Knowledge of operational readiness, service transition and BAU support models.
Experience working with document-intensive processes such as account maintenance, transfer requests, dividends, investor communications or shareholder services
Neutral 2–4 sentence summary of what working at this company is like, drawn from public reviews and press coverage. Tone, collaboration style, pace, benefits highlights.
