Modern product engineering requires the production of highly accurate digital simulations, allowing engineers to create prototypes and understand the real-world performance of materials. Currently, legacy software platforms from companies like IBM and Dassault are generally deployed, but now startups are joining the arena, powered by generative AI.
One such is Trace.Space, coming out of Riga, Latvia. This AI-driven platform has been designed for engineers to develop industrial products.
With Western manufacturing currently under enormous pressure to compete against its fast-paced Asian counterparts, the race is on to develop these platforms to speed up product development. Indeed, we recently saw the appearance of Luminary, as well as Dessia Technologies, which also has a platform to automate engineering processes through AI.
Trace.Space is, in particular, designed for the process around the development of electric and autonomous vehicles, satellites, robots, semiconductors, and medical devices. Traditional solutions tend to be based ‘on premise’, whereas Trace.Space takes a modern, cloud computing approach. It can also allow manufacturers and suppliers to collaborate on shared product requirements, thus reducing response times.
Janis Vavere, co-Founder and CEO of Trace.Space, told TechCrunch: “Every company in the world that builds complex regulated products in automotive, medical, aerospace, and so on, faces the issue that these products are becoming more and more complex, especially to design. The legacy tools and processes are struggling. IBM’s tools for this were designed in the late 80s. It’s a desktop client and needs to be installed on every computer.”
After working on Jama Software, a more modern design platform for this, he realized the need for a cloud-based, AI-driven approach: “It’s now the right moment to combine modern software architectures and UIs with AI, and apply them to these industries. Companies are looking for something better right now.”
Trace.Space is not a mere ‘AI wrapper’ he added: “We use AI models like Llama and then some deterministic AI libraries, as well as aspects of OpenAI’s LLM.”
In addition to Jama Software, where Vavere was previously a sales leader, he spent two years leading sales at the translation management software outfit Lokalise. Meanwhile, co-founder Mikus Krams previously worked in operations at Lokalise and the software development startup Chili Piper. A third co-founder, Karlis Broders, had previously implemented Jama and Polarion in large-scale projects.
Trace.Space has now raised $4 million in seed funding. Cherry Ventures led the round, joined by Riga-based Outlast Fund, along with earlier investors Nebular, Fiedler, and Change Ventures.