Profet AI launches five AI apps for semiconductor manufacturing scale-up
Profet AI unveiled Twin Accelerator, a set of five Domain Twin applications and a 30-hour apprenticeship model, to help manufacturers scale expert knowledge as semiconductor expansion strains talent and operations. Kinsus Interconnect Technology is among the early use cases, with its NPI Twin tied to broader substrate capacity expansion for NVIDIA.
Why it matters: - Semiconductor expansion is creating a skills and coordination problem as much as a capacity problem. - Profet AI is pitching a way to turn expert engineering knowledge into reusable AI workflows across plants, lines and teams. - The launch targets new product introduction, equipment service, production control, quality and workforce planning.
What happened: - Profet AI introduced Twin Accelerator with five initial Domain Twin applications. - The rollout includes 30-Hour Apprenticeship 2.0, a model for converting experienced engineers' judgment into AI capabilities. - The company unveiled the initiative in Taipei, Taiwan, on Nov. 18, 2026. - Kinsus Interconnect Technology Corp. is an early NPI Twin use case in IC substrate manufacturing.
The details: - A Profet AI survey of more than 400 manufacturing customers in Asia found plans to add more than 118 production sites or lines over the next three years. - About 88% of those planned additions are semiconductor-related. - The Semiconductor Industry Association reports more than 160 announced semiconductor ecosystem projects across 30 states since 2020, representing more than $828 billion in planned private investment. - Georgetown University's Center for Security and Emerging Technology published research in September 2026 highlighting the specialized skills needed to run new chip facilities. - Under 30-Hour Apprenticeship 2.0, experienced engineers guide task-specific AI agents using prior cases, troubleshooting methods and technical judgment. - Profet AI says the model aims to develop six to eight AI apprentices with about 30 hours of expert guidance. - Once reviewed and approved, that knowledge can help other teams assess recurring problems and decide next steps. - Higher-risk decisions still require human review. - New lessons can be captured and reused. - Kinsus' broader new product introduction transformation combines engineering records, predictive analysis and expert knowledge. - Harvard Business Review Taiwan said the broader effort improved cross-functional collaboration efficiency by 40% and cut new-employee ramp-up time by 35%. - Kinsus' NPI-Nexus initiative also received a 2026 Digital Transformation Dingge Award. - On Nov. 17, NVIDIA announced an agreement under which Kinsus plans to increase advanced ABF substrate manufacturing capacity dedicated to NVIDIA by more than 3.5 times. - Kinsus said the expansion includes more than $1 billion in NVIDIA investment support. - Production is expected to begin in the second half of 2027. - Twin Accelerator's five applications are NPI Twin, FAE Twin, PMC Twin, QE Twin and HRBP Twin. - NPI Twin uses product specifications, pilot-run data and prior engineering decisions to assess yield risks and process changes. - FAE Twin uses equipment service histories and troubleshooting knowledge to support initial fault diagnosis and technical responses. - PMC Twin links demand forecasts, inventory and supplier lead times to flag shortages and support procurement decisions. - QE Twin connects quality records, investigation histories and test data to identify similar issues and guide follow-up inspections. - HRBP Twin combines employee competencies, organizational requirements and AI-agent capabilities to identify skills gaps and recommend human-and-AI team setups. - Profet AI says Domain Twin combines AutoML for structured manufacturing data with AI Studio for building AI assistants, digital employees and workflow-connected applications. - The platform includes access controls, approvals, audit trails and on-premises deployment options. - Profet AI was founded in 2018, is headquartered in Taiwan, serves more than 400 manufacturers across Asia and has worked on more than 5,500 AI application projects. - More information is available at the company's website.
Between the lines: - The announcement positions AI as infrastructure for operational memory, not just automation. - Profet AI is emphasizing governance because semiconductor manufacturers need many AI agents to work across sensitive, high-stakes processes. - Kinsus gives the launch a concrete manufacturing example tied to a major chip-supply-chain expansion.
What's next: - Profet AI will need to show whether the five applications can move from pilot use cases to broader deployment across manufacturing sites. - Kinsus' substrate expansion is expected to ramp into production in the second half of 2027. - As semiconductor investment grows, manufacturers are likely to keep looking for tools that preserve expert know-how while scaling output.
Disclaimer: This article was produced by AGP Wire with the assistance of artificial intelligence based on original source content and has been refined to improve clarity, structure, and readability. This content is provided on an “as is” basis. While care has been taken in its preparation, it may contain inaccuracies or omissions, and readers should consult the original source and independently verify key information where appropriate. This content is for informational purposes only and does not constitute legal, financial, investment, or other professional advice.
Sign up for:
Taiwan Industry Weekly
The daily local news briefing you can trust. Every day. Subscribe now.
Check Your Email!
We sent a one-time activation link to: .
Confirm it's you by clicking the email link.
If the email is not in your inbox, check spam or try again.
Welcome back!
is already signed up. Check your inbox for updates.