The way workforce software gets delivered is changing. We're changing with it — and taking our customers with us. Here's the shift, and why it plays to exactly what we're best at.
Workforce processes — from recruitment and onboarding through to pay and invoice — are moving along a clear path. Most implementations still sit right at the start: full manual builds. We're only now stepping into "today", where the base is automated, so every project from here begins there and heads toward agentic.
2cloudnine installation and configuration was fully manual and took months to stand up. Pay and invoice runs were done largely by hand. The value was in getting the software in and configured correctly. Most existing implementations still sit here.
2cloudnine's "Basecamp" now automates the base platform setup. But it isn't hands-off — we still tailor the configuration (e.g. IR, rates, pay/bill configuration and GL) to each business, and support the transition with testing, integration, enhancements, data migration and training. Most businesses are here now.
AI agents run relevant business processes — assembling pay and invoice runs, resolving routine exceptions, escalating the genuinely novel ones — on a schedule, with people approving what matters. Even IR and rate configuration is increasingly AI-assisted, reading award rules straight from Fair Work.
Installing and base-configuring the platforms we deliver is increasingly automated — 2cloudnine's "Basecamp" is the clearest example, and Seven20 and 3B are heading the same way. Standing the software up is no longer where we add value.
To the work automation can't do alone — moving businesses off legacy systems onto the new platform, tailoring their award and IR rules, integrating, testing and training — plus the data readiness, guardrails and assurance that let automation be trusted with real pay.
Most agentic projects fail on data quality and governance, not on the AI. In payroll there is no room for "mostly right." Owning the rules and the assurance is exactly what we do.
We build on 2iq, 2cloudnine's MCP server, on Salesforce Agentforce, and on agentic capability across Seven20 and 3B — so agents can work end to end, from recruitment and onboarding through to pay and invoice.
This isn't all tomorrow. Real AI capability is shipping in the products we implement now — and we help you put it to work.
2iq is 2cloudnine's MCP server: the secure layer that lets AI agents carry out real platform actions like timesheets, pay and invoicing, under proper controls. The agents themselves get built on top of it, and 2cloudnine will keep extending it from there. nuuco is one of the partners piloting 2iq with them, so we're inside the build rather than reading the release notes afterwards.
2cloudnine now uses AI and OCR to read submitted timesheets — turning them into data automatically, cutting manual entry and errors.
3B forms can now capture a photo — a licence or passport — and use OCR and AI to extract the details automatically. No re-keying, no typos.
Generative AI builds portal pages for you — candidate apps, client portals — instead of hand-building every screen.
With 3B and workflow automation, we can wire AI into almost any process — for example, an automated candidate-interview chatbot that captures answers and scores candidates.
To be clear: none of these exist as products today. They're the targets we're designing for, chosen because they're the jobs our clients actually lose hours to every week. Each one only becomes safe once the rules, the data and the guardrails underneath it are right — which is precisely the work we do now.
Assembles the pay and invoice run on a schedule: pulls approved timesheets, applies the rules, flags what doesn't reconcile, and hands a person the exceptions to approve before any money moves.
Chases missing and mismatched timesheets ahead of the cut-off, follows up with the worker or the client directly, and escalates only what it genuinely can't resolve.
Reads award and enterprise agreement changes as they're published, drafts the configuration change, and shows a consultant exactly what it would alter before anything is applied.
Watches certificate and licence expiries across the workforce, chases renewals, and keeps each worker's compliance status current without anyone having to run a report.
Matches available, qualified workers to open shifts, offers them out in priority order, and closes the gap without a coordinator working the phone all afternoon.
Takes a client's question about an invoice line, traces it back through the timesheet and rate that produced it, and drafts the answer with the evidence attached.
We'll say plainly where each of these sits — in design, in pilot, or live — as we get there. If you'd find one of them useful, tell us: it moves up the list.
We won't recommend AI for your processes before we've used it on ours. That means being straight about where we've got to: two parts of delivery are genuinely AI-assisted today, and there's a clear list of what we're working on next. As that lands, it means less manual effort, quicker turnaround and a lower cost to serve that we intend to pass on.
Workshop recordings and notes become structured requirements, process descriptions and user stories, drafted for a consultant to check rather than written from scratch. It takes days out of the front of a project and means less gets lost between the room and the build — which matters most on the award and IR detail, where a missed nuance surfaces later as a pay error.
Our enhancement work now runs design-to-build with AI supporting the technical build. Work moves from an agreed solution design into a drafted build far more quickly, with a consultant reviewing, testing and owning whatever ships. It compresses the build without shifting the accountability, and it's most valuable on the well-understood changes that would otherwise sit in a queue behind larger work.
None of the following is in production yet. We're naming them because they're the parts of delivery that still take the most manual effort, and because we'd rather be measured against a real list than a promise of "AI-powered delivery".
We implement and tailor your front and middle office on the Salesforce platform today — and guide you to agentic when you're ready, safely. Autonomy where it's safe, humans where it matters.
What doesn't change is the hard part we've always owned: the award interpretation, the rates, the integration, the clean data. What changes is what it's for. Those foundations are now what make an AI agent trustworthy enough to run a pay cycle — and the assurance that proves the run is right before money moves.
And it comes down to people. Our depth of experience across complex payroll and workforce implementations — and the specialists who deliver them — is what makes nuuco the right partner: the domain knowledge, the credentials and the track record to deliver it successfully.
Anyone can demo an agent. Running one that gets pay right every cycle — at scale, without losing control of compliance or accuracy — is the hard part. That's the work we do.
This is the direction we're building toward — not a menu to buy today, but how we'll take you into agentic as the technology and platforms mature. It builds on the implementation and payroll foundations on our services page that make any of it safe.
A fixed-scope audit of your data, processes and governance — where agents can add value, where they can't yet, and what has to be true first.
Designing and building agents on Agentforce and MCP, surfaced in Slack — including scheduled agents for pay and invoice cycles (2cloudnine).
Human-in-the-loop checkpoints, guardrails, and QA of agent output — so an agent's pay or invoice run is verified correct before it's committed.
Running and monitoring your agents in production, tuning them as rules and volumes change, and keeping a human accountable for outcomes.
Whether you're still manual or ready to hand a pay cycle to an agent, start with a readiness conversation. We'll be honest about what's feasible now and what's coming.