The platform · AI

AI on one member record, in Australia

AI is only as good as the record underneath it. Laid over five systems, it answers faster from five versions of the truth. Nexy's AI works on one governed member record and your own knowledge: questions answered, histories summarised, risks surfaced, reports that used to take a dashboard project. Governed by your roles and moving to sovereign Australian infrastructure.

Your data, your tenant, siloedGrounded in your recordsMetered billing, published

How does Nexy use AI?

Nexy uses AI inside the platform, on your own records. Ask Nexy answers plain-language questions from your CRM, interaction histories are summarised on the member record, overdue actions and at-risk members are surfaced automatically, and AI reporting replaces traditional dashboard projects. Every client tenant is siloed, and AI inherits your roles and audit logging.

Ask Nexy

The report you need is a question.

Ask Nexy lets an administrator question the CRM in plain language: how many, how much, since when, compared with what. The answer comes from your records, in seconds, with the working shown. The dashboard project that used to stand between a question and its answer is gone.

This is what AI can do when it lives on the record instead of beside it: a bolt-on assistant can read your documents; it can never count your members.

A bolt-on chatbot is another vendor holding your knowledge.

The market is filling with AI assistants you attach to a membership system: another contract, another copy of your content on someone else's platform, another sovereignty decision made by default. They can be genuinely useful, and they all share the same ceiling: they can read what you upload, but they cannot see the membership.

Nexy's AI is baked in. It works on the same record everything else writes to, inside your tenant, behind your roles, on the audit log. That is why it can answer questions a bolt-on never will, and why the answer to “where is our data?” stays exactly one platform long.

What the AI does

Work, not novelty.

Nine things AI does inside the platform. Every one of them works on your own member record, under your own roles, on your own audit log.

Ask Nexy

Question your CRM and admin area in plain language. Counts, trends, comparisons, arrears, attendance, answered from your own records in seconds.

AI-powered reporting

Nexy is replacing traditional BI dashboards with reporting you ask for. No data warehouse, no dashboard project, no waiting on a consultant to build the view you needed yesterday.

CRM automation

Renewal sequences, follow-up, escalation and data hygiene running on rules your team can read, with AI drafting the steps and people approving them.

Histories, summarised

Open a member and read the relationship in a paragraph: the year's interactions, payments, participation and open matters, distilled.

Signals worth acting on

Members at risk, arrears building, engagement falling away, cases ageing. Raised while there is still time to do something about it.

Drafting on the record

Replies, notes and communications drafted from what the record already knows, then edited and sent by a person, with delivery visible on the member.

Answers for members

Coming to the member Hub: conversational answers grounded in your own content, deep into PDFs and policies, cited back to the source that answered.

Siloed and secured

Every client tenant is completely separated. Your AI sees your data only, under your roles, on your audit log.

Governed like everything else

AI inherits role-based access and audit logging. It reads what the person asking may read, and what it did is on the record.

For members · coming to the Hub

The end of the search bar.

Members do not want ten links; they want the answer. AI in the Hub will answer member questions from your organisation's own knowledge, deep into the PDFs, policies and resources where the answers actually live, cited back to the source. The 9pm question gets a 9pm answer, grounded in what your organisation has published.

Questions

The questions boards are actually asking.

Seventeen answers, in writing, before the first call. If your board is weighing AI, these are the questions worth putting to every vendor on the shortlist.

What it does

What does Nexy actually use AI for today?
Ask Nexy answers plain-language questions from your own CRM records. Interaction histories are summarised on the member record. Overdue actions and members at risk are surfaced before renewal season rather than discovered after it. AI reporting is replacing traditional dashboard projects. Grounded answers for members are coming to the Hub.
Does AI make decisions about our members?
No. AI on Nexy is assistive: it surfaces, summarises, searches and answers. Decisions that affect a member's standing, entitlements, money or membership remain with your people, made in the platform and written to the audit log like any other change.
Can AI sit on top of our existing systems instead?
It can, and it will answer faster from whatever those systems contain, including where they disagree. If the membership database says a member is financial, the website says lapsed and the events tool holds an old email address, an AI layer across all three gives you a confident answer built on the conflict. That is why Nexy puts AI on one governed member record rather than over several. The question worth asking any vendor is not what their AI can do, but what record it is reading.

Your data and where it runs

Why not just use Copilot or ChatGPT?
General-purpose assistants are useful, and they work across the documents and email they can reach. They do not hold your governed member record: standing, entitlements, payments, participation and history, kept current and permissioned. Nexy's AI reads that record directly, inherits the roles your team already works under, writes what it did to the audit log, and runs to requirements we publish, including that member data is never used to train a model.
Is our member data used to train AI models?
No, and that is a permanent position rather than a current setting. Member data is not used to train our models or anyone else's. It is one of the five requirements any AI provider must meet before we name them.
Where does the AI run?
AI workloads on Nexy are moving to sovereign Australian infrastructure so inference happens onshore. The requirements for that move are published on our sovereign AI page, before any provider is named, so you can hold us to them.
Can we turn AI off?
Yes. Your organisation decides whether AI capability is on at all, and what it can touch. It is something you govern, not something that happens to you.
Who can see what the AI can see?
AI inherits your role-based access controls. It reads what the person asking is already allowed to read, and nothing more. A membership officer and an events coordinator asking the same question can get different answers, because their permissions differ.
Is AI use recorded?
Yes. AI activity is governed by the same audit logging as the rest of the platform, so what was asked and what was done is on the record rather than in someone's browser history.
What happens if the AI gets something wrong?
Answers are grounded in your own records and content, so errors are traceable to source rather than invented from the open internet. Because AI is assistive rather than decisive, a wrong answer is a wrong draft or a wrong figure to check, not an action taken on a member's account.
What if your AI provider changes its terms, prices or model?
That is exactly why the requirements are published and the platform is not built around one provider. Nexy's AI capability is designed so the provider underneath can change without your processes being rebuilt around it.

Governance, cost and the board

Our data is messy. Will AI make that worse?
AI will make it visible, which is uncomfortable and useful. A tool that answers questions from your records will expose duplicates, gaps and contradictions that a report previously smoothed over. That is an argument for one governed member record, and for fixing the data as part of migration rather than after go-live.
Will AI replace our staff?
Not in any organisation we have worked with. It removes the parts of the job nobody was hired to do: reconciling systems, building a report to answer one question, reading a year of history to prepare for one conversation. The work that keeps members is the work only people do.
How is AI billed?
AI consumption is metered through our AI provider and billed as used, at the rates in your order form. It is never bundled invisibly into the licence, so you can see what it costs and decide whether it is worth it.
What should our board ask before approving AI use?
Five questions cover most of it: what problem are we solving, where does the data go, who can see what, who is accountable when it is wrong, and what happens if the vendor changes. We answer all five in writing, and this page is the short version.
How does this sit with the Privacy Act reforms?
The reforms increase scrutiny of automated decisions affecting individuals and of where training data lives. Nexy's position is deliberately conservative on both: AI is assistive rather than decisive, member data never trains a model, and workloads are moving to Australian infrastructure. We are not a substitute for your own legal advice on your obligations.
How do you keep up with how fast this is moving?
With the tools available to a software team in 2026 and beyond, we intend to stay on the cutting edge as they develop. That matters for two reasons. AI brings an onslaught of new security challenges, and staying current is how a platform holding member data stays ahead of them rather than behind. And it changes how quickly we can respond: an issue that once took a release cycle to turn around can be fixed and shipped in a fraction of the time. For a system of record, that pace is a governance property, not a novelty.

Bring a question your reports cannot answer.

Ask it in the demo, in plain language, and watch the answer come from the record. Then ask where the AI thinks, and compare our answer with anyone's. The argument behind the product is in Issue 7 of The Platform Realist.