The archive · Hardware & Devices · Product decision · 2018–2025
SixSense bets fabs swap manual inspection for no-code AI; $8.5M, 100M chips
Singapore startup SixSense sells fabs no-code AI defect detection; Peak XV's Surge led its $8.5M Series A.
SixSense
What the business is
SixSense's AI platform turns semiconductor fab data — defect images and equipment signals — into real-time defect detection, root-cause analysis, and failure prediction for chip manufacturers.
How it started
Engineers Akanksha Jagwani (CEO) and Avni Agrawal (CTO) founded SixSense in Singapore in 2018. After evaluating sectors from aviation to automotive, they picked semiconductors: fabs were full of dashboards and SPC charts that displayed data without analyzing it, leaving pattern-spotting and root-cause tracing to engineers — time-consuming, subjective, and unable to scale.
What happened
SixSense raised an $8.5M Series A led by Peak XV's Surge in July 2025, bringing total funding to about $12M, with Alpha Intelligence Capital, FEBE, and others participating. The platform runs at GlobalFoundries and JCET, has processed over 100 million chips, and is compatible with inspection equipment covering more than 60% of the global market. Customers reported up to 30% faster production cycles, a 1-2% yield boost, and a 90% cut in manual inspection work.
How it ended up
Still scaling as of the July 2025 TechCrunch report: working with fabs in Singapore, Malaysia, Taiwan, and Israel, expanding into the U.S., and riding the geopolitical tailwind of new fab construction across Southeast Asia and America.
Background
SixSense was founded in 2018 by Akanksha Jagwani and Avni Agrawal, who bet that semiconductor fabs would trade manual, fragmented quality inspection for AI-native tools that ordinary process engineers can run. After interviewing more than 50 engineers, they concluded that fabs were drowning in dashboards and SPC charts that merely displayed data, leaving anomaly-spotting and root-cause tracing to humans — slow, subjective, and unable to scale with process complexity.
The product translates raw fab data — defect images and machine signals — into early warnings, defect detection, root-cause analysis, and failure prediction. Its key wedge is that process engineers, not data scientists, can fine-tune models on their own fab data and deploy them in under two days without writing code. By July 2025 the platform was live at GlobalFoundries and JCET, had processed more than 100 million chips, and covered inspection equipment used across over 60% of the global market.
In July 2025 SixSense raised an $8.5M Series A led by Peak XV's Surge, bringing total funding to about $12 million. Customers reported up to 30% faster production cycles, a 1-2% yield improvement, and a 90% reduction in manual inspection work. The company was expanding from Singapore, Malaysia, Taiwan, and Israel into the U.S., betting that new fab construction across Southeast Asia and America — where plants start without legacy systems — would open more doors for AI-native inspection.
What has to be true
- Jagwani and Agrawal chose semiconductors only after systematically comparing sectors and interviewing 50+ engineers, so the product matched a real, manual bottleneck.
- Building for process engineers instead of data scientists removed the scarcest resource in the adoption equation and cut deployment to under two days.
- The startup picked a moment when fab expansion in Malaysia, Singapore, Vietnam, India, and the U.S. gave greenfield plants a reason to adopt AI-native tooling from day one.
- Real customer results — yield gains, faster cycles, 90% less manual inspection — gave the company credibility that raw funding rounds could not.
What can be applied
Win by removing the specialist bottleneck: hand control to the process engineer. Sell into decisions, not data, and fabs adopt without a data-science team.
Aftermath
As of the July 31, 2025 TechCrunch report, SixSense had just closed its $8.5M Series A (about $12M total) and was scaling deployments across fabs and OSATs in Singapore, Malaysia, Taiwan, and Israel while expanding into the U.S. It counts GlobalFoundries and JCET as customers, with more than 100 million chips processed through its platform. The founders framed regional fab expansion — driven by geopolitical shifts in chip manufacturing — as the tailwind that lets AI-native inspection enter plants without legacy systems.
Sources
- Female-founded semiconductor AI startup SixSense raises $8.5M
- Founded & Funded: 2025-Q3 Portfolio News
spotted an error? The archive wants to know.
Your turn
You just read one. Describe what you are building, and see who is betting on the same thing.
Free account · 3 free questions · no card