AI can make individuals more productive. But turning those gains into lasting value for an entire organisation requires something more: connected knowledge, shared ways of working and a willingness to rethink how work gets done.
Tasks that previously required considerable time can now be completed faster, and ideas can be turned into prototypes with remarkable speed - with AI, the pace of software development is changing.
But there's a difference between making individual tasks more efficient and making an entire organisation more effective.
It's a distinction we've been exploring at Logivity as we evolve our own technology and development processes.
When individual gains don't add up
Like many organisations, we saw AI being adopted across different parts of our development process. Individual productivity improved, yet the collective gains weren't as significant as expected.
Much of the value was still getting stuck between people, processes and tools.
Requirements, development and testing remained relatively separate activities. Knowledge was scattered, communication depended on handovers, and AI was being used in different ways without a common foundation.
To change that, we've been working towards a more connected approach, starting with a shared knowledge base. Not only documenting technical architecture and coding practices, but also capturing our understanding of customers, their challenges and what our solutions are intended to achieve.
Because developing software isn't simply about translating requirements into code. It's about understanding the purpose behind them.
We're also rethinking traditional roles and handovers to create a more connected process, supported by AI where it creates meaningful value.
Faster isn't necessarily better
One of our early learnings has been just how easy AI makes it to create prototypes. Ideas can be explored and customer value tested much earlier.
But when a prototype looks convincing, it's tempting to assume that a finished solution is just around the corner.
It rarely is.
Building something quickly doesn't replace the need to understand whether it should be built, how it fits into the wider architecture or what is required to make it reliable at scale.
And as development accelerates, testing must evolve alongside it. Otherwise, faster production simply moves the bottleneck further down the process.
There's also a difference between creating something new and transforming existing systems and architecture. The latter requires patience, consistency and a longer-term perspective.
A humantic approach: humans and agents, together
As we've progressed, one conclusion has become increasingly clear: scaling AI isn't simply about giving agents more capabilities. It's about enabling humans and agents to work effectively together.
We've started calling this a humantic approach.
It means designing processes, responsibilities and communication around the strengths of both humans and AI agents, rather than assuming automation alone will deliver the desired results.
And communication works both ways. Just as agents need to understand human instructions, people need to understand what agents are doing and how their contributions fit into the bigger picture.
The objective isn't to make everything agent-driven. It's to make the whole process work better through human–agent collaboration.
Interestingly, this challenge has much in common with the transport logistics industry itself.
A single transport can involve multiple companies, systems and responsibilities. Planning, operational performance, costs and emissions are often managed separately, even though they are closely connected.
Improving one part of the operation doesn't automatically improve the whole.
At Logivity, our purpose is to connect the different parties, data and processes involved in transport logistics - making it easier to collaborate, use available capacity more effectively and improve both economic and environmental performance.
We see a similar principle in our own AI journey.
The greatest opportunity isn't necessarily in making every individual activity faster. It's in connecting activities so that improvements in one part create value across the whole.
That's how individual productivity can become scalable business value.
These reflections are based on a presentation by Jessica Öhrblad, CEO of Logivity, at Gothenburg Tech Week 2026, where she shared some of our early experiences and learnings from evolving our development approach.
Author: Logivity
