AECOM

Applied Ai Engineer

Manchester, LEVEL 4, gb

Job description

AECOM is seeking an Applied AI Engineer to join our Applied AI team, contributing to transformative AI-enabled delivery across our clients' projects. As part of AECOM's commitment to delivering sustainable and innovative solutions, the Applied AI Engineer will play a critical role in designing, building and scaling practical AI solutions that unlock measurable business value across infrastructure, environmental, and urban development sectors. 

Reporting to the Applied AI Lead, this role will focus on building agentic workflows, integrating AI with client data and systems, and turning proven solutions into repeatable offerings for many different clients. The ideal candidate will bring expertise in large language models, software engineering and data integration, with a passion for solving complex challenges in the built environment through simple, effective and well-tested solutions. 

Key Responsibilities: 

Design, build and deploy AI-enabled solutions that address real client and project problems, from discovery and proof of concept through to production. 

Deploy and operate AI solutions on Microsoft Azure, including model hosting, search and retrieval services, serverless and container compute, secrets management, identity and private networking, in line with AECOM's cloud governance. 

Develop agentic and multi-step workflows, including orchestration and multi-agent patterns, deciding where deterministic code, AI models and human review each belong. 

Build and maintain Model Context Protocol (MCP) servers and tool interfaces that give AI models secure, least-privilege access to data and systems. 

Integrate diverse data sources (APIs, databases, document repositories, scheduled exports and streaming feeds) into reliable, tested pipelines that adapt to different client environments. 

Host, configure and optimise local and private large language models for clients whose data cannot leave their environment, including model selection, quantisation, inference serving and hardware sizing. 

Build retrieval-augmented generation and document-intelligence pipelines, including OCR, extraction, classification and cited answers, over large and unstructured document sets. 

Optimise token usage, latency and cost through model selection, prompt caching, batching and incremental processing, working within defined usage budgets. 

Establish evaluation, testing and monitoring practices for AI solutions, including evaluation datasets, regression tests, run logs and accuracy measurement on real data. 

Design data models, dashboards and reports that present AI outputs clearly to project teams and decision-makers. 

Apply security, data protection and responsible AI practices, including defences against prompt injection, data residency controls and human-in-the-loop review, ensuring alignment with AECOM's global standards. 

Turn successful solutions into reusable, configurable components that can be deployed for multiple clients with minimal rework. 

Document work through specifications, implementation plans and post-delivery reviews, capturing evidence of the value delivered. 

Support bids and proposals by shaping AI solutions to client requirements, writing technical responses, estimating effort and running cost, and preparing demonstrations for client presentations and interviews. 

Mentor colleagues and support AI training programmes, fostering a culture of innovation and safe, effective AI adoption. 

Present work outputs to both technical and non-technical audiences, translating complex AI concepts, and their limitations, into clear, layman's terms.

Meet AECOM

Explore company profile

Filters

Create your account

Create an account to save jobs and track applications.