TechnologyOne’s PLUS Moment
What Councils Need to Know To Gear Up for Agentic AI
Last week, TechnologyOne unveiled PLUS, its new “agentic enterprise AI” platform1.
The launch was billed as a watershed moment for TechnologyOne and its customers, marking the introduction of a new intelligence layer that connects finance, HR, assets, and property data to surface insights, anticipate needs, and accelerate decisions.
For many in local government, that sounds like the breakthrough they’ve been waiting for. A smarter, faster, more connected back office. The promise is compelling, and what’s emerging in PLUS reflects the first practical form of agentic capability. That is, a system that can anticipate needs, surface relevant data, and complete process actions inside the ERP.
On the agentic AI maturity curve I wrote about a few months ago2, I’d place it somewhere between Level 1 (The AI Intern) and Level 2 (The AI Assistant): guided, context-aware, and task-capable, but still operating within defined system boundaries. It’s progress, just not yet orchestration.
The term “agentic” is everywhere in AI marketing right now, though most implementations are still early in that journey. True agentic systems act across workflows with a degree of autonomy and purpose. Others, including TechnologyOne’s PLUS, operate in the formative stages of that curve, where intelligence assists rather than orchestrates. Understanding that distinction helps set the right expectations and gives a clearer view of what TechnologyOne is offering its customers today.
PLUS is best understood as a large-language-model enhancement layer that brings conversational capability and guided execution to the TechnologyOne environment. It primarily lets users ask questions in plain language, retrieve structured information, and generate reports, summaries, or comparisons. Increasingly, it will complete routine actions such as approvals, requisitions, and updates. It’s designed to be intuitive and genuinely useful, turning ERP data into contextual responses and process outcomes.
That makes PLUS a conversational and assistive system rather than an autonomous one. From company material and early demonstrations, it doesn’t yet act across modules or execute workflows beyond defined boundaries, nor does it make independent decisions. “Human-in-the-loop” guardrails confirm its purpose as a safe, bounded layer for insight and task execution. In practical terms, it’s a smart, guided interface that helps users act more efficiently within the TechnologyOne system they already use.
That distinction matters for government leaders. A large-language-model layer can enhance how staff interact with data and complete everyday tasks, yet it doesn’t by itself reshape how the organisation operates. For that, a platform must progress to higher levels of the agentic maturity curve. At those stages the systems act, learn, and coordinate across domains rather than simply assist within them. Right now, PLUS sits between the first and second levels of that curve. It is capable, helpful, and safe, but not yet orchestrative.
This matters for councils because most still operate in fragmented data environments. Finance, assets, planning, people, and customer systems rarely align, whether the environment is best-of-breed or a single vendor suite. What separates maturity from disorder isn’t the number of systems or the promise of a single platform, but the capability to connect and properly govern them .
Integration discipline, not consolidation, determines whether data flows cleanly or ends up in spreadsheets and workarounds. The golden rule is actually pretty simple: integration capability, not system count, defines digital maturity.
A genuinely agentic system capable of reasoning across functional silos also relies on consistent data models, clear integration pathways, and mature governance. These are precisely the areas where most councils struggle today, not necessarily through neglect but because of structural realities that have built up over years like skill gaps, system changes and constrained investments.
No AI can fix those underlying problems. And a large-language model can’t reason its way through poor structure or inconsistent data. So if the foundations aren’t ready, the AI won’t clean the mess. Fragmentation, data quality issues, and broken integrations don’t disappear under an intelligence layer. They just surface faster (or not at all if the query times out).
Commercially, TechnologyOne’s strength has always been its positioning around simplicity. The company has carefully cultivated the idea of one vendor, one stack, and one update cycle. It is a message that has proven powerful in the local government market because it speaks directly to councils’ appetite for predictability, accountability, and local support. As a marketing narrative, it has been remarkably effective.
I think that positioning, though effective, sits uneasily with where enterprise architecture is heading in general but definitely in relation to AI. Agentic systems thrive on interoperability, not control. They work best when they can move freely across boundaries, connecting to platforms like Microsoft 365, Salesforce, or ServiceNow where critical parts of council’s operational activity either lives or is moving. That trend has long been visible beneath the surface, but it’s now unmistakable in the growing number of public tenders focused on remediation and transformation at the customer layer. And often through platforms outside their incumbent ERP providers.
That does not diminish TechnologyOne’s role as custodian of core ERP data. In fact it remains central to council operations. But as customer engagement increasingly takes place across other platforms, its irrefutable that their contribution must sit within a wider, more interconnected ecosystem.
With the release of PLUS, TechnologyOne now occupies the early stages between Level 1 and Level 2 on the agentic maturity curve. It can anticipate needs, surface data, and complete routine actions inside the ERP. That progress gives councils a safer, more guided way to interact with their information. For most customers, that’s the right place to be because it is secure, familiar, and focused on improving confidence before moving toward more open, orchestrated systems.
But to climb higher and become a genuine agentic platform, it will need the ability to open up. That means not just integrating with other systems, but federating intelligence across them so reasoning and context can flow between platforms. It could be as “easy” as partnering. At least, that’s the direction the leading agentic ecosystems are already taking.
And that’s the paradox. The same Power of One philosophy that built TechnologyOne’s success also constrains how far PLUS can evolve as an enterprise-wide agentic platform. Its strength in unification becomes a limitation when intelligence needs to operate beyond its own ERP dataverse. Remaining closed keeps them comfortable and successful at Levels 1-2. Becoming open could make them 10x the company they are today.
For now, and for most councils, PLUS will feel like progress. It will give staff a new, intuitive way to interact with data. One that doesn’t require new training, new governance, or new integrations. It will let managers query information in plain English and get instant answers.
It looks modern, it probably feels intelligent, and it can’t really break anything. Security has been at the core of the solution. That’s exactly why it will be attractive to their base. PLUS will deliver the appearance of transformation without the disruption that real transformation requires.
It will give leaders an AI initiative to announce. It is something tangible that staff can use, executives can demonstrate, and auditors can understand. But it won’t change how the organisation actually works. That’s a positive because local government really does not do change well.
In relation to data, the real challenge lies not in the presence of silos but in the quality of what sits inside them and the council’s ability to manage and connect that data across boundaries. In that sense, PLUS is a new way to see your data, not a new way to run your organisation, and that’s not a flaw. It’s an honest reflection of where most councils are today, and it remains valuable so long as everyone understands where this AI’s impact begins and ends.
Councils that mistake PLUS for an AI architecture shift risk overestimating its impact and underinvesting in the groundwork that true transformation requires. That is not good for anyone. The irony is that the work needed most (things like improving data quality, cleaning and reconciling information across systems, and building cross-platform governance) can already be accelerated through other forms of AI.
The real opportunity is to do both. PLUS provides a safer, more accessible way for staff to interact with data, while specialised AI tools can strengthen that data by profiling, cleansing, and mapping it so future intelligence layers have something trustworthy to work with. The councils that succeed will be those whose vendors, like TechnologyOne, help them mature along that path. Not just by adding visibility, but by enabling readiness. Ignoring that step risks building new visibility on top of old problems.
I think we can all agree that AI progress doesn’t happen all at once. Like every other technology adoption cycle we’ve lived through, it moves in stages. And that goes for the organisations buying the technology and the vendors selling it. As a technology analyst, I receive around half a dozen press releases and briefings each week from software companies entering the AI conversation. It has been non stop in this space for more than 12 months. A few recent examples help illustrate how differently each one is approaching the journey.
Epicor launched its AI solution earlier this year based on IBM’s Watson X platform. Their model combines a vertical AI layer for manufacturing and supply chain with a broader platform architecture capable of orchestrating multiple domains.
Dayforce have taken a different route, embedding small AI agents directly into HR workflows. It doesn’t claim to be agentic, because it isn’t. It’s a feature enhancement focused on measurable efficiency. That honesty keeps it grounded.
The big guns including ServiceNow, Salesforce, Microsoft and even Accenture and Google Cloud have gone further still, building hybrid agentic layers that operate (and govern) across multiple products and vendors. These are genuine AI platforms of interaction, not just enhancements or systems of record.
By comparison, TechnologyOne’s current positioning feels closer to Dayforce than to the platform models of Epicor, ServiceNow, Salesforce or Microsoft. That’s not a weakness, it’s simply a stage. What matters is context. Council buyers need to understand which stage they’re stepping into, and whether it aligns with where they intend to go next.
So let’s bring this all together.
Before councils start planning around PLUS, it’s worth pausing to understand what’s actually being offered. Where does this new intelligence live? Some elements sit inside existing TechnologyOne modules, automating approvals, updates, and workflow actions, while others sit above them in a conversational layer that interprets, retrieves, and surfaces relevant data. That distinction matters. If it’s embedded, you’re buying automation. If it’s layered, you’re buying visibility. Both have value, but they’re not the same thing.
It’s also worth asking what “agentic” really means in this context. Can PLUS act across systems, or does it simply answer questions within TechnologyOne’s own data schema? And how open is the platform? Can it exchange data with Microsoft, ServiceNow, or anything beyond its own walls? And if so, how?
Then look inward. If your council data is inconsistent or your integrations incomplete, PLUS won’t solve that. That’s your job. But it will make those issues more visible. So be sure to build that in.
Finally, consider the commercial model. With AI there is usually price and cost. Pricing is typically the annual licensing fee (not all vendors charge one). Then there is cost3 based on “interactions” or “conversations” with the LLM. Because PLUS appears to be a private LLM environment rather than one that calls public models like ChatGPT, Claude, or Gemini, there are unlikely to be separate token costs. That said, the detail I have access to is not clear on this point. Just read the footnote for more context to get started on your thinking.
At the end of the day, TechnologyOne’s AI investment is genuine and strategically significant. If I was a client I would want it because it absolutely has the potential to make daily life inside its ERP suite easier.
But PLUS is not yet the intelligent operating core implied by the word “agentic” in the media. At least not by whatelse I have seen and experienced in the market. For now, it’s best viewed as the next step in the evolution of the ERP interface. A great conversational layer that sits above structured data, powered by a large language model. That makes it interesting, but also something to approach like a proof of concept4 rather than a finished platform.
For councils, that’s still progress. It’s safe, auditable, and consistent with existing governance expectations. Just don’t mistake it for the architectural leap that companies like Microsoft, Salesforce, and ServiceNow are already building toward. A full AI platform where agents don’t just explain your data, they autonomously act on it.
AI really does matter. Don’t be afraid of it. But also, the real opportunity for local government isn’t to buy the loudest product. It is to build the conditions under which any of them can succeed. That means cleansing and classifying data, simplifying integrations, reducing system complexity, and establishing clear governance around data sharing and access. Once that groundwork is in place, choosing an AI layer, whether PLUS or anything else, becomes a question of fit, not faith.
As at launch, PLUS is a promising step in both usability and guided execution. It is a sign that the ERP interface is changing to how people will increasingly work. But it’s not yet a transformation in architecture. And that’s perfectly fine, provided everyone understands the difference. Because in local government, AI won’t replace data governance. It will simply expose the absence of it.
https://itwire.com/it-industry-news/market/technologyone-unleashes-game-changing-ai-revolution-in-enterprise-software.html
https://www.afr.com/technology/ai-is-disrupting-software-so-techone-built-its-own-chatgpt-20251008-p5n0uy
This curve originates from Councilio’s Agentic Maturity Model first outlined in the August 2025 article “Have You Tried Our New AI Agent?.” It describes six stages, from The Intern to The Autonomous Partner, mapping the evolution from reactive AI to truly agentic systems.
At the heart of the cost question is a new challenge: what do we actually measure in the AI era? If MIPS and FLOPS defined computing performance in the mainframe age, today that role is played by tokens. That is the unit of work for most large language models. Each token represents roughly a word or part of a word. Every time you prompt a model like ChatGPT, the total number of tokens processed (input and output) determines the cost. Providers such as OpenAI, Anthropic, Google, Mistral, and Meta publish per-token prices, but comparing them isn’t straightforward. Don’t think about it as list prices so much as runtime cost-to-serve, which depends on assumptions about token volume, throughput, and response behaviour. That’s the real “cost” of each prompt-response interaction. If the model runs entirely inside the vendor’s own environment, there are no token costs. Those only appear when the system starts calling out to public models.
Some councils are already experimenting with their own versions. Devonport, for example, has built an internal assistant using Microsoft’s AI tools with speech-to-text capability.