Planacy's internal AI transformation lays the foundation for a smarter SaaS product

Planacy's AI formula: experience builds better products

Planacy's development team has experimented with AI tools for several years. In 2025, they reached a tipping point: the tools had matured enough to be applied at scale in day-to-day work.

Since then, Planacy has undergone a full internal AI transformation – delivering significant productivity gains and, just as importantly, deep organisational learning. This internal shift, where AI is embedded in core functions, is the foundation for building a truly AI-powered product.

“To understand where AI creates real customer value, we first needed to make the transformation ourselves. We believe experience builds better products. That meant fully adopting AI internally before introducing AI functionality for our customers.” – Mikael Edh, CPO at Planacy

How AI transformed product development at Planacy

Planacy’s development team uses the agentic software Augment Code to automate the majority of code production. Developers collaborate on cases with custom built agents in much the same way as with human colleagues.

This agentic development model has fundamentally changed how teams work. Developers no longer spend their time writing code. Instead, they focus on defining intent, setting direction and reviewing output before release.

“Today, a developer together with AI-agents can take on a case, identify the best solution, test it and deliver release‑ready code in a matter of hours instead of days.” – Mikael Edh, CPO at Planacy

The product team also uses AI to collaborate on requirement documents and release notes, as well as help analysing how upcoming releases will impact existing customer configurations. This enables earlier insight into communication needs and potential risks.

While AI has significantly accelerated product management work, the productivity gains are most dramatic in development. With AI producing most of the code, the product team has now become the primary bottleneck in the development process.

“Our development productivity is so high that we need to grow our product team to keep pace. Previously, one product owner would generate work to about five developers. With AI, that model no longer works – we need to rethink our organisational structure.” – Mikael Edh, CPO at Planacy

As a result, product management and development roles are converging. Both functions must understand each other’s domains to succeed in an AI‑driven environment. Developers need deeper product and customer insight to evaluate AI‑generated code against business objectives. Product managers, in turn, must understand how the product is built to write specifications that AI agents can interpret and execute correctly.

The challenge: no room for uncertainty when accuracy is critical

Getting the organisation on board with the AI transformation was relatively straightforward. The benefits were visible from day one. Still, legitimate concerns remain.

AI can make mistakes, and in some cases, hallucinate. For Planacy’s customers, who rely on the platform for financial planning, errors can have serious consequences. Accuracy is non‑negotiable. That is why quality always takes precedence over speed or productivity.

“We need AI tools that both our team and our customers can trust. That’s why systematic security checks and human review are built into our processes before any AI‑generated code is approved.” – Mikael Edh, CPO at Planacy

The results: increased productivity and an AI‑native mindset

The productivity gains in product development are clear. But equally important is the shift in how the organisation thinks and works. Planacy has developed a genuinely AI‑native mindset – a critical capability as they help customers adopt AI in their own financial planning processes. With a faster development pace, the team is confident in its ability to innovate and deliver increased customer value.

Beyond coding: High‑speed, high‑quality content production with AI

With development teams spending less time on manual coding, Nimblr also explored new AI use cases beyond engineering. One result is a new internal tool supporting content creation.

Nimblr’s security awareness courses and personalised training content are continuously updated to stay relevant and localised across regions and languages. During the first quarter alone, the new AI‑powered content engine enabled one person to produce the same volume of content that previously required three people plus a professional translator for an entire year.

The tool includes a translation memory tailored specifically to security awareness, ensuring both consistency and accuracy. Productivity has increased dramatically, while content quality has improved.

What’s next: supporting customers on their AI journey

Planacy is currently developing and testing AI functionality scheduled for release later this year. Over time, these capabilities will help customers transform the way they work – just as Planacy has done internally.

The AI‑powered platform will enable businesses to extract deeper insights from their financial data. Examples include explaining changes in key KPIs, generating reports and presentations, and identifying errors before they impact financial statements – all within the SaaS platform customers already use.

“We could have launched AI features much earlier. Instead, we chose to move deliberately. Our focus is on delivering safe, meaningful functionality – not AI for AI’s sake. We want to guide our customers into the AI era, not overwhelm them.” – Mikael Edh, CPO at Planacy

Advice to SaaS companies starting their AI journey

Based on Planacy’s experience, Mikael shares three practical lessons for SaaS leaders:

1. Start small. Begin with a clearly defined use case. A small project is better than no project at all, and it allows you to test models and ways of working with low risk.
2. Test, fail and test again. AI evolves rapidly. A tool that didn’t work three months ago may now be ready. Revisit and reassess continuously.
3. Standardise to scale learning. Use a shared AI toolbox across the organisation. This allows teams to train agents collectively and maximise the return on the tools you choose.