TL;DR:
- SixSense, a Singapore-based semiconductor AI startup, has raised $8.5 million in Series A funding, bringing its total to $12 million.
- Peak XV’s Surge, Alpha Intelligence Capital, and FEBE Ventures joined the round.
- The startup provides real-time chip defect detection to help fabs improve yield and reduce manual inspection.
- SixSense’s platform already powers fabs like GlobalFoundries and JCET.
AI for Real-Time Chip Defect Detection
SixSense has developed a platform that enables semiconductor manufacturers to detect chip defects and predict failures using real-time data from production lines. The Series A funding of $8.5 million was led by Peak XV’s Surge (formerly Sequoia India & SEA), with backing from Alpha Intelligence Capital and FEBE Ventures. This brings the startup’s total funding to around $12 million.
Founded in 2018 by engineers Akanksha Jagwani (CTO) and Avni Agarwal (CEO), the company is targeting a core issue in chipmaking — the inability to convert massive volumes of production data into actionable real-time insights. Their AI platform analyzes defect imagery and sensor data to drive automated root cause analysis and yield optimization.
Data Intelligence Missing in the Fab
Despite the semiconductor industry’s reputation for precision, co-founder Agarwal told TechCrunch that inspection processes in fabs are still largely manual and fragmented. After speaking to more than 50 engineers, the founders discovered that SPC charts and inspection dashboards show data but do little to interpret or act on it.
“The burden of using data still falls on engineers,” said Agarwal. “Spotting patterns and tracing anomalies is subjective and doesn’t scale with complexity.”
Akanksha previously built automation solutions for Hyundai Motors and GE, while Agarwal engineered analytics systems at Visa, some of which are now protected trade secrets.
Making AI Easy for Process Engineers
SixSense’s platform stands out because it doesn’t require users to be data scientists. Process engineers can deploy and tune AI models within two days using fab data — with no coding required.
The platform supports:
- Defect classification and anomaly detection
- Root cause analysis
- Failure prediction models
SixSense competes with in-house AI teams using tools like Cognex and Halcon, along with AI-driven inspection solutions from startups like Landing.ai and Robovision.
The Data
| Metric | Value |
| Total Funding | $12M (source) |
| Series A Round | $8.5M (source) |
| Chips Processed | 100M+ (source) |
| Manual Inspection Reduction | 90% (source) |
| Yield Improvement | 1–2% (source) |
| Supported Equipment Coverage | 60%+ of global market (source) |
Global Scaling Fueled by Geopolitics
SixSense already works with global fabs including GlobalFoundries and JCET, helping reduce production cycles by up to 30% and significantly improving yield.
Its primary clients include:
- Foundries
- OSATs (Outsourced Semiconductor Assembly and Test)
- IDMs (Integrated Device Manufacturers)
The company is now expanding into the U.S., following strong uptake in Singapore, Malaysia, Taiwan, and Israel.
Geopolitical shifts — especially the U.S.-China chip decoupling — have created new opportunities. As U.S. CHIPS Act funding encourages domestic manufacturing, SixSense is gaining traction with greenfield fabs not tied down by legacy software systems.
“New fabs are more open to AI-native approaches like ours,” said Agarwal.
