ISLAMABAD: Pakistani robotics data startup World Context hosted its first hackathon at the University of California, Berkeley, bringing together 13 teams to build projects using the company’s datasets as part of an effort to expand artificial intelligence and technology ties between Pakistan and the United States, it said on Sunday.

Founded in Pakistan in January, World Context records how workers perform physical tasks in factories and other real-world settings and converts the recordings into datasets used to train robots.

A hackathon is a short, intensive event in which teams collaborate to develop and demonstrate technology projects around a particular problem or set of data.

“The talent on display at Berkeley showed how advanced the industry is in the US, and that there are tremendous opportunities to learn from and partner with US institutions and companies,” World Context founder and CEO Sherdil Ali Khan said in a statement.

“We can bring these learnings back to Pakistan to improve the state of technology there as well.”

The event, held in partnership with student organization Blockchain at Berkeley, was sponsored by Google Gemini, ElevenLabs, The Hackathon Company, Lovable, Praxis Robotics and UC Berkeley Startup, according to World Context.

Participants competed across hardware, software and visualization tracks using datasets supplied by the Pakistani company, with a total of $4,500 in cash prizes awarded.

Sortbot, a project to develop robots capable of sorting different types of objects, won the hardware track, while Reconstructors won the software category with a project focused on recreating World Context data in the physical world.

A simulation produced by the same team won the visualization track.

World Context collects what is known as human demonstration data, recordings of people carrying out physical tasks that can be used to teach artificial intelligence systems how those tasks are performed in real-world environments.

The company uses proprietary sensors, wearable gloves and head-mounted cameras to record workers performing industrial tasks, capturing information including hand and finger movements and the pressure applied during repetitive work.

It says it is collecting data from around 8,000 workers in Pakistan across textile, automotive, logistics and other manufacturing sectors and selling datasets to robotics companies in the United States and China.

Khan has previously said Pakistan’s large manufacturing workforce and reliance on skilled manual production could allow the country to carve out a role in the global robotics industry by supplying the real-world data needed to train machines.

Manufacturing accounts for 14.8 percent of employment in Pakistan and 20.7 percent of urban employment, according to the Pakistan Labour Force Survey 2024-25.

World Context said UC Berkeley was selected for its first hackathon because of the university’s strengths in robotics and engineering and the opportunity to work with its students and researchers.

Khan said the projects developed during the event showed students were able to use the company’s human demonstration datasets in ways World Context itself had not anticipated.

The company said it planned to work with universities on similar events in the future as it explores closer collaboration with academic institutions.