How Lánluas Consulting Used AI to Simplify Migration, Automate Invoicing, and Transform Operations

We’d just come out of a three-month project to implement Productive, moving away from our previous Salesforce and Kantata platform. We decided to split the functionality between HubSpot for sales and Productive for running our core business operations.

Interestingly, AI wasn’t the reason we chose Productive. We selected a plan that included AI because there were other features we needed. At the time, we hadn’t really anticipated how valuable those AI capabilities would become.

What surprised us was just how much easier AI made the entire transition.

We ended up using Productive AI throughout the implementation, from data migration and cleansing to platform configuration, troubleshooting, and training. It wasn’t just answering questions about how the system worked. Because AI could work with our actual data and configuration, it could identify things we’d set up incorrectly and help us fix them.

Essentially, we were getting two answers at once: Here’s how you do something in Productive, and here’s what’s wrong with how you’ve currently set it up. Would you like me to fix it?

That combination made an enormous difference.

We managed the entire implementation in-house rather than bringing in an external implementation partner, largely because of the AI capabilities in Productive.

For a project of that complexity, that was a significant benefit we hadn’t expected when we first chose the platform.

Migrating data between enterprise systems is never straightforward, particularly when you’re dealing with financial information, resource schedules, bookings, and projects. It’s not simply a matter of moving a few line items from one spreadsheet to another.

We had a fairly complex migration process. We migrated our sales data from Salesforce to HubSpot, then used the native integration to import customer and project information into Productive. On top of that, we had to migrate our resource bookings and scheduling data, including plans extending up to a year ahead.

That data gets complicated very quickly. You have bookings measured in hours or days, different working arrangements, weekends, public holidays, and all sorts of other variables to account for. We used CSV extracts from our previous system and gave that information to Productive AI, asking it to create the corresponding projects, bookings, and associated details.

We could ask AI to reconcile the imported information against our original data and generate reports showing whether everything matched. That meant we could verify that the resource bookings we’d planned months in advance had been accurately transferred into Productive.

We also used AI extensively for data cleansing. We could ask it to identify deals with incorrect values or currencies and update them automatically. In the past, that might have meant manually opening and correcting 50 individual deals.

Even seemingly small things, like correcting naming conventions, formatting, or inconsistent data, became much easier to manage. Those little cleanup tasks can take an extraordinary amount of time during a migration. Being able to hand them over to AI was phenomenal.

One of the most useful aspects of Productive AI during our implementation was its ability to run in the sandbox environment.

When you’re migrating critical business data, there’s naturally some hesitation about letting AI make changes. Some of our people were understandably nervous about trusting it to update information.

The sandbox gave us a way to address that. We could import our data, ask AI to make changes, and then review the results without affecting our production environment. If something wasn’t quite right, we’d refine the instructions, run them again, and see whether we got the desired outcome. Once we were satisfied, we could apply those same instructions in production.

That was particularly valuable for the more complicated data cleansing and configuration tasks, where you want to be absolutely confident that the changes are correct. Rather than asking the team to simply trust AI, we could demonstrate that it worked.

Being able to test, review, and refine AI instructions before deploying them made the whole process much more comfortable for our team.

And ultimately, it helped us move through the implementation much more efficiently.

One of the biggest areas for us now is invoicing. We’ve built AI agents to manage the invoicing process using the timesheet data recorded in Productive.

We operate on weekly timesheets, so the agent takes that information, processes it, generates the invoices, and creates the individual line items. We’ve also been able to configure the agent to present the invoice information in the format we need.

That might sound relatively straightforward, but our invoicing arrangements aren’t all the same. Some projects are prepaid, others are postpaid, and we have different arrangements involving hours or days purchased in advance.

The agent has to take those differences into account when processing the information.
We then use Productive’s integration with Xero to handle the invoicing workflow and bring payment information back into the system.

Our first run of the invoicing agent was successful, and the figures balanced upon completing our manual reconciliation. That was a really encouraging result, particularly because we’d only just gone live with the new platform.

We’re now in a position where AI can take on a meaningful part of that administrative workload rather than having our operations team manually prepare all that information.

Absolutely. Alongside AI, we’ve made good use of Productive’s automatic time-tracking capabilities.

Around 70% of our staff use automatically generated timesheets. Because we schedule resource reservations in Productive, the system can automatically create timesheet entries based on those bookings. Consultants then review and verify their timesheets at the end of the week.

Instead of having to remember and enter every hour they’ve worked, they’re essentially managing their timesheets on an exception basis.

For other team members whose work is more fragmented, we use the Outlook calendar integration to help them record time in smaller increments, such as 15-minute or hourly blocks.

To be clear, we’re not using AI to generate the timesheets themselves. That’s handled by Productive’s existing functionality. Where AI comes in is the next stage, taking that recorded time and turning it into invoicing data.

It’s the combination of automation and AI that makes the process so effective. Our consultants spend less time on routine administration, and the information flows much more easily into the financial side of the business.

This has probably been one of the most important changes for us.

Previously, we operated with a fairly closed approach to business information. People didn’t necessarily have access to the data they needed, and getting answers often meant going through our operations team.

As part of our move to Productive, we deliberately made more of that information available to our executive team, account managers, and salespeople. But giving people access to more data is only part of the solution. You also need to make it easy for them to understand and use it.

That’s where AI has made such a difference. Our teams can now ask questions directly about the information in Productive.

For example, they can ask when a particular consultant is available or what impact taking on another project would have on resource availability. Rather than relying on someone to manually review schedules and compile an answer, they can get that information directly from the system.

We’re also using AI to build reports, generate pulse reports, and share information with the wider team.

That’s especially important for us because we operate across Australia, New Zealand, and the UK. We’re effectively a 24/7 business, so we need information to be available when people need it, regardless of their location or time zone.

We’ve also extended access to our external contractor pool, allowing contractors to update their availability so we can incorporate it into our resource planning.

We’ve moved from a legacy, somewhat paper-based way of working to real-time bookings, real-time visibility, and immediate answers.

It’s a significant improvement in how we operate.

We’ve introduced a very simple principle across the business: Ask AI first.

When someone has a question about Productive, our training encourages them to go to AI before approaching the operations team. That might be a question about how to use a particular feature, how something has been configured, or how to find information in the system.

Because Productive AI understands the platform and can work with our actual business data, it can provide much more useful answers than a traditional help resource on its own. This was valuable during the implementation, but it’s become just as important now that we’re live.

We have people using Productive who weren’t involved in the migration or initial configuration. They’re coming into a new system, learning new processes, and figuring out where everything lives. Instead of organizing additional training sessions or continually asking operations for help, they can simply ask AI.

The impact on our operations team has been substantial.

Their workload has dropped significantly because employees are finding answers themselves rather than directing every question to operations.

Of course, there’s an important prerequisite to making that work. You need to invest in getting your data right.

If you want employees to rely on AI for business information, the underlying data needs to be accurate and properly maintained. That requires a commitment from the organization.

But once you’ve got that foundation in place, AI becomes a natural part of everyday operations. For us, encouraging people to go to AI first has been incredibly successful.

The biggest surprise was how little external support we needed.

We’d undertaken a fairly substantial project, migrating multiple systems, bringing across complex data, configuring a new platform, and introducing new processes across the business.

Normally, you’d expect considerable back-and-forth with an implementation partner or support team. Throughout the deployment, we only needed to contact Productive’s support team twice for clarification.

Everything else was largely handled through the AI capabilities available to us. Whether we were trying to understand a feature, troubleshoot a configuration issue, correct migrated data, or figure out how to accomplish something, AI was there to help.

And the two questions we did raise with Productive’s team were answered very quickly.
Looking back, the combination of the platform itself and its AI capabilities made the implementation experience exceptionally good.

Considering we hadn’t originally selected Productive because of AI, it’s remarkable how central it became to the success of our deployment.

Now that we’ve completed the migration and are building up more live operational data, we’re focused on continuing to improve how we use the platform.

One of our next priorities is scenario planning for deals while they’re still in the presales stage.

We want to model different project scenarios and understand how they would affect our future resource requirements and forecasts.

At the moment, some of that forecasting is still split between our sales system and Productive. We’d like to bring more of it together so we can make decisions with a clearer picture of what’s ahead. We’re also looking at adopting resource requests more extensively, giving our operations and delivery leaders a more structured way to manage competing resource demands.

And as we continue building out our reporting, we’ll have more opportunities to make business information accessible through AI.

For me, the real value of AI isn’t simply that it can answer questions or automate individual tasks. It helps people work more independently, reduces the burden on operations, and makes the information we already have much easier to use.
We started this journey looking for a better platform to run our business. What we discovered was that AI could help us implement the platform, get our data right, and fundamentally improve how our teams interact with it.

That’s been the unexpected benefit, and one we’re looking forward to building on.