Access Construction, the construction division of the UK software group The Access Group, announced on 15 October that it has brought what it calls construction-specific agentic AI across its estimating, project management and ERP products. The release, carried on GlobeNewswire with a London and New York dateline and syndicated in full here, describes the newest generation of Access Evo for construction as extending “traditional systems of record into a connected system of action.”
The products underneath it are the three Access Construction already sells: Access ConQuest for estimating and tendering, Access Fonn for site and project delivery, and Access Coins for finance, commercial, operations and workforce management. The release says the platform and its enhanced Access Evo capabilities are available starting on the day of announcement across “key markets, including the UK, Ireland, the United States, and Australia”, and the company’s boilerplate puts Access Construction at more than 130,000 users globally across construction, specialty trades, real estate and infrastructure, with The Access Group headquartered in Loughborough.
The Four Named Capabilities
The release is specific about what shipped, which makes it checkable in a way that many AI-layer announcements are not:
- Access Evo Navigator — a role-based personalised workspace pulling information, tasks, priorities, communications and actions from connected applications into one view
- Access Evo Researcher — a business-connected AI research experience for asking questions in everyday language, investigating information, analysing reports and documents, and preparing content
- Access Evo Assistants — role-based AI assistants that complete specialist and recurring tasks, work in the background and prepare outputs, “with human oversight and approval where required”
- Access Evo Workflow — a visual drag-and-drop builder for connecting data that lives in multiple systems, in place of exporting, copying and consolidating by hand
The release describes the assistants as operating on strict “Human-in-the-Loop” guardrails, carrying out complex background tasks “from extracting data to drafting compliance forms, while keeping human experts in total control of final approvals,” and says no technical knowledge or coding is needed.
On governance, the release states that Access Evo holds ISO 27001 certification for information security, that customer data remains within the customer’s secure Access environment and is never used to train external AI models, and that existing permissions decide what each person can reach — the AI can only access data the user’s software permissions already allow.
That last point is doing more work than it looks. A layer that reads across estimating, site and finance is, by construction, a layer that can surface a margin position to somebody whose job does not include seeing it. Inheriting the host system’s permission model rather than building a second one is the correct answer, and it is the specific risk that an AI layer over an ERP introduces. This newsroom has covered the same question from the opposite end: when a Sydney builder’s job-management software opened itself to AI agents over MCP, the notable part of the documentation was that it told builders their own site diaries are untrusted text an attacker can write into.
Evo Is Not New. Its Construction Generation Is.
The word “newest” in the release is load-bearing, and the timeline is worth giving readers rather than leaving to inference.
Access Evo is a group-wide programme, not a construction product. The Access Group debuted it in its HR software — the group’s own April 2025 announcement says it “debuted Access Evo within the HR profession in June last year,” which places it in June 2024 — and that same announcement, dated 4 April 2025, reported that the group had “surpassed one million users on Access Evo, its new AI-enabled software experience,” alongside a plan from chief executive Chris Bayne to “release over 50 products enhanced with responsible AI” over the following twelve months.
Construction was already inside that programme before this week. Access Construction’s own product pages describe Access Evo as “the secure, construction-specific AI layer that sits across your Access Construction suite,” and the group’s US construction pages market the ERP as Access Coins Evo. What 15 October adds, on the release’s account, is the four named capabilities above and the extension across all three specialist products rather than the ERP alone.
This is not a discrepancy and it is not presented as one — the company’s own wording is “the newest generation.” It matters for a different reason: it tells you what kind of product decision this is. An AI layer built once at group level and rolled out division by division — the April 2025 announcement describes 50+ further products planned that year “across multiple industry sectors, including healthcare, finance, legal, recruitment, hospitality, not-for-profit, education and manufacturing” — is a fundamentally different bet from an AI product built for contractors by people who only sell to contractors, and the trade-off is the whole story of this announcement.
The Superlative Is the Company’s, and It Is the Claim to Watch
Alex Boury, General Manager of Access Construction, is quoted at length, and the sentence that will travel is this one:
“With Access Evo, Access Construction is now the only provider that connects context across the entire construction lifecycle.”
That is an exclusivity claim about a crowded market, made by the vendor, and it is reported here as the vendor’s claim rather than as a finding. The release does not define what “connects context across the entire construction lifecycle” would require, and this publication has not attempted to adjudicate it; readers should treat it as positioning.
What can be said about the competitive ground is narrower, and what follows is this publication’s reading rather than a finding. The suite Boury is describing — estimating through to finance, under one owner, with one AI layer over it — is unusual among construction-native vendors, many of which hold one or two of those three positions. It is much less unusual among the horizontal ERP groups, and it looks like the same consolidation logic that has driven this year’s deal flow: Nemetschek’s acquisition of HCSS and Hexagon’s $535m agreement for Rocscience both bought adjacency rather than capability.
The rest of Boury’s quote is the more defensible version of the argument: that construction firms “have invested heavily in software, but too often the value is limited to what people can find, manage or do within individual applications,” and that the opportunity “is much bigger than adding another AI tool or making individual tasks faster.” On the survey evidence, that is the right read. Bluebeam’s 2027 AEC Technology Outlook, published two days earlier, found that what carried firms from AI pilot to production was integration with existing systems and tools, named by 47% of those that got there — ahead of data quality and collaboration.
Who Else Is Named
Two external voices appear in the release.
Aly Pinder, given as Research VP, Aftermarket Services Strategies at IDC, comments on the category requirement: that as contractors move “from reactive record-keeping to automated, proactive execution without compromising compliance or control, it is crucial to embed agentic capabilities directly across core operational and financial systems while maintaining strict enterprise security and role-based data governance.” An IDC speaker listing gives his full title as Research Vice President, Worldwide Aftermarket Services Strategies.
Steve Higgins, Construction Director at Ascia Construction, supplies the customer voice, saying the firm is “very excited by the potential of Access Evo to make a real difference to our daily working lives” and that harnessing AI agents to automate routine tasks “could further reduce administration and give our people more time to focus on what they do best”. Ascia describes itself as one of the leading independent construction companies “in Hampshire and across the South of England”, working across commercial, industrial and defence, education, health, leisure, refurbishment and residential sectors from a base on the Hampshire–Sussex border. Higgins’s biography on the company’s own site says he joined in 2022 as Construction Manager, became Construction Director in 2023, and has 30 years of experience.
Worth noting for what the reference is: Ascia is an existing Access customer, the subject of a published Access Group case study on its use of Access Fonn in which Higgins is also the quoted executive. His quote in this release is forward-looking — potential, could — rather than a report of results, which is the honest form for a capability shipping on the day of announcement.
What the Release Does Not Cover
The release names no pricing or packaging for the new capabilities, and Access’s own construction pricing pages quote on request rather than publishing rates. The UK pricing page for Access Coins Evo, the ERP underneath the layer, carries “Request Pricing” in place of every figure; what it does publish is the shape of a deal — “Pricing is per user”, a choice of “annual, quarterly, or monthly billing”, and a first contract that “covers implementation costs, with a standard 3-year initial term”. Access Fonn is sold on a different basis again — “a flat fee with unlimited users, projects, and subcontractor access” across four tiers — which is worth knowing before reading any per-seat assumption into the layer that now spans both. Whether Evo’s new capabilities are included in existing licences or priced as an upgrade is therefore not established by any published Access material this publication could find, and a prospective buyer should ask.
Nor does the release carry adoption or performance numbers for Evo in construction specifically: no count of construction customers on the Evo layer, no time-saving or margin figures, and no named deployment beyond Ascia’s prospective one. The 130,000-user figure is for Access Construction overall, not for Evo, and the one-million-plus Evo user figure is group-wide and eighteen months old.
What It Says About the Shape of the Market
Analysis, labelled as this publication’s own.
The interesting thing about this launch is not whether the four capabilities work. It is that the horizontal suite vendors have now arrived at the same product thesis as the construction-native startups, from the other direction, and on the startups’ own chosen ground.
Across the past six weeks this newsroom has covered Kuro raising €10m to build AI for general contractors on twelve customers, six of them tier-one DACH firms, Volve putting an orchestration layer over tender documents and Trunk Tools launching Cortex to read a contractor’s drawings and connect them to the rest of the project record. Every one of those is a bet that the value sits in the connective layer above a contractor’s systems rather than in another system. Access’s answer is that it already owns the systems — estimating, site and finance, under one roof, serving 130,000 users — so the connective layer is a feature it can ship rather than a company someone has to build.
Both propositions can be true at once, and which wins is an empirical question about depth. A group that builds one AI layer for every division it owns gets distribution and governance for free and pays for it in domain specificity; a startup that builds for contractors only gets the specificity and has to buy the distribution. What this release does establish is that construction software buyers evaluating AI are no longer choosing between an AI product and their incumbent. Increasingly, the incumbent is the AI product — and on Bluebeam’s numbers, the firms that got AI into production were the ones that found somewhere to put it.
