A Singapore -based deep tech startup is called six sense An AI-powered platform has developed which helps semiconductor manufacturers to predict and detect potential chip defects on production lines in real time.
It has raised $ 8.5 million in Series A, causing its total funding to about $ 12 million. The era was led by Peak XV’s Serge (East Sex India and C), with participation from Alpha Intelligence Capital, Fibe and others.
In 2018, the establishment of engineers by Akanka Jagwani (CTO) and Avni Aggarwal (CEO), the six -year is aimed at addressing a fundamental challenge in semiconductor manufacturing: changing raw production data, defect from images to equipment signs, helping in real -time insight to prevent quality issues and improve yields.
Regardless of the sheer volume of data generated on the fab floor, what was for co-founders was a surprising lack of real-time intelligence.
Akanksha makes a deep understanding of manufacturing, quality control, and software automation through its experience, which creates a solution of automation for manufacturers such as Hyundai Motors and Ge and LED product development in startups such as ambibe. Aggarwal adds technical experiences in visas from his time, where he created a large -scale data analytics system, some of which were later preserved as trade secrets. A skilled coder with a strong background in mathematics, he was interested in implementing AI in traditional industries beyond Fintech for a long time.
Together, the pair evaluated areas ranging from aviation to motor vehicle before landing on a semiconductor. Aggarwal told Techcrunch that despite the reputation of the semiconductor industry for accuracy, inspection procedures are largely manual and fragmented. He said that after talking with more than 50 engineers, it became clear how the quality check is done is an important place to modernize it.
Agarwal stated that Fabs are filled with dashboard, SPC charts and inline inspection systems today, but only display data without further analysis. “The burden of using it to make decisions still falls on engineers: (them) should check the spot pattern, discrepancies, and find out the root causes. It is time -consuming, subjective, and the growing process is not well scale with complexity.”
Techcrunch event
San francisco
,
27-29 October, 2025
Sixmen provided to engineers with initial warnings to address potential issues, before they proceed with capabilities such as detecting defects, the basic cause analysis and prediction of failure.
Aggarwal said that the platform of Sixasense is specially designed to use by process engineers rather than data scientists. “Process engineers can fix the model using their own FAB data, deploy them within two days, and rely on the results-without writing a line of all codes. That is what makes the stage powerful and practical.”
The competitive landscape includes in-house engineering teams, using devices such as Cognx and Halllon, integrates the inspection equipment manufacturer AI into its system, and startups including Landing.AI and Robovijan.
The AI platform of Sixes’ is already in use in major semicondrose manufacturers such as globalfoundry and JCET, with more than 100 million chips to be processed to date. The founders said that customers have reported a 30% faster production cycle, an increase of 1-2% in yield and 90% decrease in manual inspection work. The system is compatible with inspection devices that cover more than 60% of the global market.
Aggarwal said, “Our targeted customers are large-scale chipmakers-including foundry, outsourced semiconductor assembly and test provider (OSATS), and integrated device manufacturer (IDM).” “We are already working with Fabs in Singapore, Malaysia, Taiwan and Israel, and are now expanding in America”
Geophysical stress, especially between the US and China, where chips are made, is taking shape again, running new manufacturing investments around the world.
“We are seeing that Malaysia, Singapore, Vietnam, India and America expands aggressively-and it’s a tailwind for us. Why are we already based in this area, and many of these new features are starting fresh-weighing them without heritage systems.
