Major programmes arrive on target and on budget when they are civil- and structurally-led. Twenty-five years of proof across nuclear, defence, marine and energy — now multiplied by AI, automation and the full 2D–10D BIM stack.
"Most major projects are most likely to arrive on target and within budget when they are civil- and structurally-led — because the ground, the structure and the sequence are where programmes succeed." JASON JAMES McCAFFREY · PROJECT DIRECTOR, AES LTD
The next generation of nuclear will be won on a modular kit of parts — factory-built, site-assembled, and certain on programme. Those are exactly the disciplines AES has spent twenty-five years delivering: safety-case rigour, DfMA-led construction, 4D-driven schedule certainty, and assurance that stands up to the regulator.
From first appraisal to final certificate — calculations, drawings, reports and director-level delivery advice across the whole project lifecycle.
Steel, concrete, masonry and timber — calculations, drawings and specifications ready for Building Control.
→ INFRASTRUCTURERoads, external works, drainage, SuDS and flood-risk — the ground works that make the building possible.
→ EXISTING STRUCTURESCondition and subsidence surveys, defect diagnosis and remediation design you can act on.
→ GROUNDDesk studies to 50-metre marine piles — investigation scoped, interpreted and turned into a foundation that works.
→ BUILDABILITYSequences, propping and buildability designed with the build, not after it.
→ COMMERCIALNEC3/4 administration, tender reporting, procurement and site monitoring through to completion.
→AI services that turn spend, carbon and risk into decisions — built and run with an engineer's eye for delivery, grounded in your own data.
AI that cleans, classifies and connects fragmented spend data so the value hiding in the tail becomes obvious and stays visible.
→AI SERVICEModel Scope 3 across the supply chain, baseline it accurately, and track every reduction against target, not estimate.
→AI SERVICEContinuous due diligence and risk scoring that surfaces critical exposure before it ever reaches the programme.
→AI SERVICEKey terms, dates and obligations held in one place, with renewals and breaches flagged early enough to act on.
→AI SERVICEVisibility into category cost drivers and indices, so negotiation starts from evidence instead of guesswork.
→AI SERVICEResource, progress and compliance tracked live, so slippage is found early, when it is still cost-effective to fix.
→AI SERVICEAn agentic layer that reads your documents, extracts what matters, and answers in plain language, grounded in your own sources.
→Twenty-five years of civil, structural and architectural engineering across the sectors where getting it wrong is not an option — nuclear, defence, energy, marine and petrochemical.
Design, analysis and assessment of structures built to survive the real world, from process plant to nuclear-grade facilities.
→SITE · MARINE · LINEARSite, marine and linear infrastructure: earthworks, drainage, foundations and the groundwork everything else stands on.
→CONCEPT → CONSTRUCTIONFunctional, buildable space, concept to construction, with delivery kept in mind from the first line.
→REGULATED · HIGH-CONSEQUENCERegulated, high-consequence programmes delivered to the standard those environments demand.
→PROCESS · ONSHOREProcess, marine and onshore infrastructure for large-scale energy and petrochemical programmes.
→BS 6349 · OFFSHOREPiling, moorings and coastal works, including ocean-energy arrays and offshore foundations.
→2D–7D are the established, contractually-referenced core. 8D–10D — safety, lean construction and industrialised / prefabricated delivery — extend BIM across the whole lifecycle; powerful, but not yet fully standardised, so each is specified to the project.
Drawings, plans, sections and schedules — the contractual record every model is checked back to.
→3DOne federated model. Clash-checked geometry every discipline builds against.
→4DThe programme linked to the model — sequence, phasing and progress you can see.
→5DQuantities and cost drawn from the model itself, updating as the design moves.
→6DCarbon, energy and lifecycle performance modelled early, while change is cost-effective to fix.
→7DAsset and maintenance data built in for handover — a model that keeps working.
→8DHazards, exclusion zones and method risk designed in before anyone is on site.
→9DWaste, bottlenecks and rework stripped out — every step earning its place.
→10DOff-site manufacture, DfMA and prefabrication driven straight from the model.
→Using Autodesk’s official Revit Public MCP Server for Revit 2027, AI reads and understands the live model directly: detecting which projects are open, querying elements by category, family, level, type and parameter, pulling full element data, then selecting, zooming and opening the right views — and exporting drawings, images and schedules on command.
Read-only and fully human-controlled today, with create-and-modify capability on the roadmap through a separate write server. Paired with Dynamo, Autodesk Construction Cloud and Forma, the model stops being a static deliverable and becomes something you can question — and act on — at the speed of conversation.
Twenty-five years of EPCI engineering leadership — now fused with the AI, automation and BIM stack built and run in-house today.
The field hasn't settled on one title. Here are the ten in current use, from those who direct AI to those who build it — each mapped to where engineering already fits.
The AI Generalist is the orchestrator. They apply AI across functions — strategy, delivery, operations, communication — rather than burying themselves in a single specialism.
→ROLE 02 / 10The most recognised builder title. The AI Engineer designs and ships AI-powered systems — taking a model from a clever demo to something that holds under real load.
→ROLE 03 / 10The AI Application Developer builds the things people actually open and use — apps and tools powered by language models, aimed squarely at an end user.
→ROLE 04 / 10The AI Automation Engineer turns repetitive, judgement-light tasks into pipelines — combining AI with the unglamorous plumbing of scripts, triggers and APIs.
→ROLE 05 / 10The AI Integration Engineer connects AI to the systems people already use — through APIs, MCP servers and the existing toolchain — so capability reaches where the work happens.
→ROLE 06 / 10The Agent Engineer builds and directs autonomous agents — systems that plan, use tools, and carry a task across several steps rather than returning a single reply.
→ROLE 07 / 10The Prompt Engineer designs the instructions and system prompts that shape model behaviour — the craft of getting reliable, on-brief output from a fixed model.
→ROLE 08 / 10The AI Solutions Architect designs the whole system — which models, which tools, which infrastructure, which data flows — and, just as importantly, what to leave out.
→ROLE 09 / 10The AI Product Engineer — often 'forward-deployed' — sits between a real problem and a working AI system, building bespoke solutions close to the people who'll use them.
→ROLE 10 / 10The Applied AI Engineer — sometimes Generative AI Developer — is the broad, practical builder: someone who puts generative models to work on real tasks and ships the result.
→If your programme needs delivery certainty, safety-case rigour and a modern digital backbone in the same team — let's talk.
