Strong software engineering skills in Python, with production experience of APIs, services and data pipelines.
Strong experience working with databases and large, complex datasets.
Good understanding of how large language model agents use tools and what makes a tool easy or hard for an agent to use.
Experience testing or evaluating large language model systems for accuracy, reliability and cost.
Practical knowledge of token efficiency in large language model systems.
Experience with AWS or a comparable cloud, containers and infrastructure as code.
Understanding of identity and access control, including OAuth 2.0, single sign-on and least privilege.
A disciplined approach to security, privacy and safe access to production systems.
Ability to work with engineers, operators and managers to understand their data needs.
Fully committed to AI, using AI tools daily and keeping pace with the latest models and agent capabilities.
Responsibilities
Map company data sources with their owners, including their structure, quality, sensitivity and access permissions.
Build pipelines that turn raw fab data into agent-ready data, including tool logs, recipes, statistical process control, metrology, test and inspection data.
Build and own connectors that give AI agents access to company data and systems using the Model Context Protocol, including the Manufacturing Execution System data warehouse, Microsoft 365, engineering file shares, tool and process data, and business systems.
Design connections that agents can use effectively and efficiently, making them easy to find and understand, economical in tokens and cost, and suitable for longer-running, multi-step and event-driven work.
Test connections with real agents on real company tasks, measure accuracy and cost, and monitor them in production without degrading source systems.
Secure connections so people can access only data they are entitled to, protect personal, commercially sensitive and intellectual-property data, guard against risks such as prompt injection, and maintain full audit trails.
Work with production, engineering, quality and business teams, IT, data owners and vendors to turn real needs into documented connections that deliver measurable gains.
Desired Qualifications
Experience building Model Context Protocol servers or tools for AI agents, and experience with Amazon Bedrock or Anthropic Claude models and agent tooling.
Experience with semiconductor, photonics or advanced manufacturing data, Manufacturing Execution Systems, statistical process control, yield or equipment data.
Experience with big data platforms and tools, such as data lakes, data warehouses and distributed processing.
Experience with event-driven systems and observability tools.
Experience in a secure, intellectual-property-sensitive environment, including Data Protection Impact Assessments and information security reviews.
We are seeking an AI Data & Integration Engineer to connect Plessey’s data and systems to AI. The company operates an advanced compound semiconductor fab and develops complex MicroLED display and optical interconnect technologies. This role will give AI agents secure, efficient and reliable access to data from across the business, including manufacturing, engineering, quality and business systems, so they can deliver real outcomes.
The successful candidate will combine strong data and software engineering skills with a deep understanding of how AI agents use tools and data. They will build the connections, shape the data so agents use it well, and prove through testing that agents can complete real work with it.
Success in this role will be demonstrated by AI agents that answer questions and complete tasks accurately and efficiently using Plessey’s data. This will include clear evidence of faster access to information, less time spent gathering and analysing data, and measurable productivity gains across fab operations, engineering and business functions.
The role reports to the Lead AI Solutions Engineer and works closely with IT, data owners and teams across the business. As AI agents take on longer-running and more complex work, this role will make sure Plessey’s data is ready for them: secure, well structured and efficient to use.
Map Plessey’s data sources with their owners: structure, quality, sensitivity and who should see what.
Build pipelines that turn raw fab data into agent-ready data: tool logs, recipes, statistical process control, metrology, test and inspection data.
Build and own the connectors that give AI agents access to company data and systems, using the Model Context Protocol (MCP). These include the Manufacturing Execution System (MES) data warehouse, Microsoft 365, engineering file shares, tool and process data, and business systems.
Design connections that agents can use effectively and efficiently: easy to find and understand, economical with tokens and cost, and ready for longer-running, multi-step and event-driven work.
Prove every connection works by testing it with real agents on real Plessey tasks, measuring accuracy and cost, and monitoring it in production without degrading source systems.
Secure every connection so people only reach data they’re already entitled to. Protect personal, commercially sensitive and IP data, guard against AI-specific risks such as prompt injection, and keep full audit trails.
Work with production, engineering, quality and business teams, IT, data owners and vendors to turn real needs into documented connections that deliver measurable gains.
Fully committed to AI: uses AI tools every day, keeps pace with the latest models and agent capabilities, and believes in a future where AI is part of everything we do.
Strong software engineering in Python, with production experience of APIs, services and data pipelines.
Strong experience working with databases and large, complex datasets.
Good understanding of how LLM agents use tools, and what makes a tool easy or hard for an agent to use.
Experience testing or evaluating LLM systems for accuracy, reliability and cost.
Practical knowledge of token efficiency in LLM systems.
Experience with AWS (or a comparable cloud), containers and infrastructure as code.
Understanding of identity and access control: OAuth 2.0, single sign-on, least privilege.
A disciplined approach to security, privacy and safe access to production systems.
Ability to work with engineers, operators and managers to understand what they need from their data.
Advantage
Experience building Model Context Protocol servers or tools for AI agents. Experience with Amazon Bedrock or Anthropic’s Claude models and agent tooling.
Experience with semiconductor, photonics or advanced manufacturing data, Manufacturing Execution Systems, statistical process control, yield or equipment data.
Experience with big data platforms and tools, such as data lakes, data warehouses and distributed processing.
Experience with event-driven systems and observability tools.
Experience in a secure, IP-sensitive environment, including Data Protection Impact Assessments and information security reviews.