Using AI to help make sense of the jumbled mess that requirement documents cause
The field of engineering, especially in complex projects like nuclear power plants and marine vessels, depends heavily on the information gleaned from written specifications, documents, and text-based data.
However, when it comes to ploughing through hundreds of pages of technical text for relevant information and data, ambiguities can arise and much time is wasted. Not to mention how tedious and laborious this task can become, which may further drain resources.
COO Suvi Ellilä noticed this pain point in the engineering industry and set out with CEO Tuomas Ritola to develop a SaaS platform optimized for technical specification analysis. Developed from research projects at the Helsinki University of Technology which started 10 years ago, the technology is now known under the brand name Selko. The MVP is currently being piloted with several major Finnish companies.
Selko examines text-based data and extracts the most relevant information that the user or professional may need to perform their work. The software combines AI, machine learning, and expertise in engineering fields such as oil and gas, energy, construction, and aerospace. It is capable of going through document data and arranging and classifying all of it, by means of intelligent technical specification classification and analysis.
The technology leaves no room for error, can be trained to automate your processes, match existing requirement models, and has the potential to significantly accelerate the compliance processes.
Users can also export Selko’s findings in a more readable, storable format. Ellilä explains that this sort of information should be stored in digital form, and not in people’s heads. “Essentially, it means that smart people can spend their day actually solving problems instead of doing manual tasks.”
A further attraction is Selko’s ease of use. The software provides all the benefits of machine learning, with no need for AI infrastructure setup or technical knowledge