HSS ProService ‘Uber-ifies’ with functional programming and agentic AI

No ‘checkout’ endpoint in hiring

Building services has long been a “notepad-and-paper” industry. For HSS to move away from this required more than just a website refresh. 

The move to a digital-only model was only possible by solving the industry’s most difficult problem. Namely, the non-linear nature of construction hire. Unlike standard e-commerce, where a transaction ends at the checkout, a hire journey is dynamic.

“You don’t sell one item at one point in time,” said Daniele Turo, CTO of HSS ProService. “You take something on site, you deliver it, then after a few days or weeks or years, you go pick it up.

“You might have issues along the way. You might also have issues with the various suppliers. You need also to have all the compliance associated with these items that you are delivering. So, it’s a much more difficult product life cycle than in a standard e-commerce setting.”

Construction plans change constantly. A digger hired for three days might be needed for two weeks, triggering different rate structures. To handle this without manual reconciliation or “surprise invoices”, Turo’s team developed what they describe as a “self-healing” finance system.

Built as part of a system – internally known as “Brenda” – the architecture tracks “daily revenue actuals”. Every day, the system looks at the full lifecycle of every contract. If a hire duration is extended, the system automatically recalculates the financial position for the entire duration of the contract in real time.

“We have this self-healing finance system that tells us every day precisely what the customer will pay us and what we need to pay our suppliers,” Turo added.

Underpinned by functional programming

To achieve this level of reliability in a “messy” real-world environment, Turo chose to use the Scala programming language, supported by the Cats library for functional programming in the back end. 

Turo opted for functional programming, which is often the choice in high-concurrency fintech environments, based on the need for “local reasoning” to provide the ability to test and change specific elements of code without causing a ripple effect across the entire distributed system.

“The power of pure functional, particularly in this very messy setting, is that you can reason very locally about your elements of code,” said Turo. “Scala itself was built as a scalable language for big systems. With these extra layers, in Cats, you have pure functional, which makes it even simpler to reason about things, which makes it easier to maintain.”

The stack is built on standard Postgres databases but relies heavily on Kafka for event streaming and OpenSearch for discovery. The entire infrastructure is headless and API-first, allowing HSS to integrate with suppliers who may still be using old-fashioned legacy systems.

From trucks to AI agents

The shift to a tech-first model has also paved the way for AI integration. Because much of the construction industry remains digitally immature, HSS uses AI to bridge the gap between unstructured communication and its structured API-based core.

The company is currently developing an agentic AI architecture that can ingest emails from suppliers – such as proof of delivery or collection – and automatically turn them into API calls.

“Instead of asking any technical effort from not very tech-savvy businesses, we are shifting our model,” said Turo. “That unstructured data today, it’s now possible to turn it into nice, structured data that we push into our system via APIs. Nothing changes at the core.”

A new blueprint for the sector

The transformation of HSS represents a total pivot in corporate identity. 

By offloading its physical kit, HSS has repositioned itself as a data-driven broker. The platform now feeds demand data back to small suppliers, allowing them to buy and maintain relevant equipment based on real-time search trends from customers that range from enterprise to consumer scale.

The HSS story serves as a case study in using cloud-native, functional architecture to bypass legacy technical debt. By treating the “hardest layer” – the financial exceptions and real-world logistics – as the starting point, the company has built a platform it hopes will help it expand into new verticals such as waste, fuel and cleaning.

Turo said: “It was a long journey because, as usual, things are much more difficult from a human process perspective. It wasn’t the lack of the technology. It’s simply that you need to take people on the journey.”