É«É«À²

Aalto University master’s students Younes Talili (School of Business) and Hien Ta (School of Science) recently completed a Customized Student Business Project with Hilti Finland. The project explored the viability of introducing Generative AI (GenAI) tools into Hilti's firestop selection process to improve efficiency and uncover future development opportunities. Ultimately, the students aimed to provide Hilti with a Proof-of-Concept for applying AI to technical tasks, alongside strategic recommendations for the company's future AI development.

Bridging commercial AI with technical engineering 

During the initial phase of the project, the team conducted extensive research and testing commercially available AI tools. They sought to understand both the capabilities and limitations of these tools, as well as the methods required to customize them for a technical domain. The students engaged with Hilti stakeholders at every step to map current processes, pinpoint user needs, and test prototypes.
Leveraging this technical and business insight, the students developed two bespoke AI agents designed to solve key workflow bottle necks for Hilti's engineering and sales teams.

Beyond delivering technical solutions, the team shaped Hilti Finland’s broader AI adoption strategy. They established guidelines for developing effective AI agents and identified key digital infrastructure improvements necessary to integrate currently available AI tools successfully into daily workflows. Additionally, the team advised on future AI opportunities and strategies to overcome the data privacy and quality concerns that typically hinder widespread AI adoption.

Key learnings and project findings

On the technical side, the project gave the students valuable hands-on experience in customizing commercial GenAI models for specific industry domains, particularly through agent building. Working closely with the client also honed their skills in stakeholder communication and technical problem-solving within real-world business constraints.

The project yielded two primary findings regarding AI integration:
Firstly, development is human- and process-centric: Creating early phase GenAI tools requires thorough understanding of business processes, alongside close collaboration with the domain experts currently involved in those processes. It is very human centric work requiring less deep technical expertise than expected.

Secondly, user education is critical to implementation: Guiding and educating end-users on how GenAI tools work, and how to use them effectively is paramount. Even an AI agent that works perfectly in testing can fail in practice if users lack adequate guidance on the tools' function and role.

Great learning opportunity and real-world results

The students’ work was well received by Hilti. For the students, the project provided an excellent overview and hands-on experience with the practical possibilities and challenges of AI adoption.

Hien Ta says that he gained experience in applying what he learned in his technical coursework to business contexts. ‘It was extremely gratifying to work on a real-world problem and help the clients. I am very thankful for the team at Hilti for giving us ample support, which was indispensable for delivering a satisfactory project.’ 

‘During the project, we had many opportunities to collaborate with Hilti's team. I especially appreciated how genuinely interested and excited everyone was about the topic, bringing new ideas to every meeting and looking for real, actionable results. Since so many companies are currently exploring the possibilities and challenges of AI adoption, the insights I gained will undoubtedly be valuable in the future,’ Younes Talili says.

According to the project's academic advisor, Roope Nyqvist, D.Sc., this showcases how two top students can successfully develop agentic AI solutions for real-world use. ‘If implemented well, the solution could recoup its development costs within a year, with full-scale deployment delivering value exceeding the lifetime cost of a development team.’

‘I’m really pleased with the results. It would have taken us two years to achieve what the students accomplished in four months. With AI, we have the potential to serve our customers faster and with higher quality,’ praised Markus Suominen, Head of Engineering at Hilti.

‘I’m also delighted with what we’ve achieved. The student team delivered exceptional work, and we now have a much clearer understanding of how AI works, how to craft effective AI prompts, and how to apply it in our work. The first two AI agents the students created already help us move forward with our development,’ commented Juho Keitaanniemi, Engineering Manager at Hilti, following the project delivery.