Sam Pasupalak, Co-Founder and CEO of Skyfall AI, has developed SKY-001 to generate engineering designs using computer-aided design (CAD) and building information modelling (BIM).
An Indian entrepreneur has introduced an artificial intelligence model designed for complex engineering tasks, claiming it can match or outperform leading AI systems at a significantly lower cost.
According to Pasupalak, the model is faster, more affordable and more accurate than some of the leading AI systems tested by his team. The model was developed over 18 months, and the founder shared benchmark results on X to demonstrate its performance.
SKY-001 aims to address a key challenge in AI-generated engineering designs: ensuring that complex designs include all the components needed for construction. Instead of generating an entire design at once, the model breaks the task into smaller stages.
For complex products, it first identifies assemblies and individual components. It then determines how each part should be constructed using visual representations, sketches and step-by-step instructions. For building design, the model uses BIM to divide a building outline into individual units and rooms.
Skyfall AI says this approach helps SKY-001 manage projects involving hundreds or thousands of steps while maintaining consistency between individual components and the overall design.
Pasupalak claims SKY-001 is between 51 and 125 times faster than the frontier models tested. He also reports that it is between 150 and 210 times cheaper, depending on the comparison.
In the CAD cost comparison shared by the founder, SKY-001 costs $0.005 per sample, compared with $0.75 for GPT-6 Astra. This represents a reported cost difference of 150 times.
The accuracy gap is much smaller. SKY-001 recorded an Intersection over Union (IoU) score of 79.7%, compared with 79.1% for GPT-6 Astra. IoU measures how closely a generated design overlaps with a reference design.
However, these results are based on Skyfall AI’s own tests and have not been independently verified. They apply to the specific engineering tasks evaluated and do not establish that SKY-001 outperforms general-purpose AI models across other applications.
Pasupalak, originally from Bhubaneswar, Odisha, is now based in San Francisco, US. He earned a bachelor’s degree in mathematics and computer science from the University of Waterloo in Canada in 2011. He later completed a master’s degree in political science and government at the London School of Economics in 2019. His studies at Waterloo included artificial intelligence, robotics and human-computer interaction.
Before founding Skyfall AI in 2025, Pasupalak co-founded AI research startup Maluuba in 2011 and served as its CEO. Microsoft acquired Maluuba in 2017. At the startup, he led researchers and engineers working on artificial intelligence and language understanding.
Also read: Viksit Workforce for a Viksit Bharat
Do Follow: The Mainstream LinkedIn | The Mainstream Facebook | The Mainstream Youtube | The Mainstream Twitter
About us:
The Mainstream is a premier platform delivering the latest updates and informed perspectives across the technology business and cyber landscape. Built on research-driven, thought leadership and original intellectual property, The Mainstream also curates summits & conferences that convene decision makers to explore how technology reshapes industries and leadership. With a growing presence in India and globally across the Middle East, Africa, ASEAN, the USA, the UK and Australia, The Mainstream carries a vision to bring the latest happenings and insights to 8.2 billion people and to place technology at the centre of conversation for leaders navigating the future.