DeepMentor is a Taiwan-based startup specializing in AI model-to-chip transformation and edge AI deployment acceleration tools. With a series of successful national and enterprise-level projects across sectors such as AIoT, medical imaging, and semiconductor AI IP design, DeepMentor became one of the winners in the IC Taiwan Grand Challenge (ICTGC) Batch 3, AI Core Technologies and Chips category.

AI Model Sizes Delay Deployment

Interest and development in AI have grown rapidly in recent years, but deployment has seemingly stalled for many use cases. This is especially true of large language model (LLM) deployments, which rely on high-performance GPUs for training and inference. As a result, deploying LLM is prohibitively costly due to the high hardware and energy costs; deployment to edge devices also becomes nearly impossible as they generally have limited space, computing power, and energy storage. Utilizing cloud AI is a common workaround, but this method might not be possible for users with high security requirements, such as factories, hospitals, or government installations. DeepMentor democratizes AI deployment by simplifying the processes of model training, compression, and hardware integration, enabling cost-effective and secure on-device intelligence.

Introducing IC Taiwan Grand Challenge

To strengthen Taiwan's position as a global semiconductor powerhouse, the National Science and Technology Council (NSTC) organizes the IC Taiwan Grand Challenge. Started in 2024, the competition originally focused purely on innovative IC designs and advanced application solutions. Today, ICTGC also honors technologies that can support and contribute to the semiconductor industry and Taiwan's industry in general. ICTGC Batch 3 emphasizes key fields such as AI Core Technologies & Chips, Smart Mobility, Smart Manufacturing, Smart Medtech, and Sustainability.

AI Model Size Miniaturization for Secure, Real-Time Edge Intelligence

DeepMentor was founded in 2018 by a team of experts with over a decade of theoretical research and practical experience in Edge computing, IC design, and system development. Specializing in AI model-to-chip transformation and edge AI deployment acceleration tools, DeepMentor's team comprises veterans from leading global and Taiwanese technology giants, including Realtek, ZyXEL, Novatek, TSMC, and Samsung.

DeepMentor's core technologies include AI Model Miniaturization, AI Algorithm Hardening, and Complete Edge AI Toolchain. Their AI Model Miniaturization technology allows 90% AI model size reduction while maintaining over 99% output accuracy and enabling efficient deployment of large AI models on resource-constrained edge devices. AI Algorithm Hardening is DeepMentor's unique and patented technology to automatically convert AI algorithms directly into chip circuits for immediate simulation and debugging on various FPGAs, drastically shortening design timelines from 12 months to just 2-3 months. DeepMentor also provides an exclusive, integrated AI pipeline for on-premise AI inference: data preparation, pre-labeling, labeling, fine-tuning, and deployment, which simplifies the development process for enterprises.

DeepMentor's solutions are uniquely positioned for strong local integration within Taiwan's robust technology ecosystem, enabling the creation and deployment of advanced AI solutions that can address global challenges. DeepMentor has also developed close relationships with local academic institutions in Taiwan, including the Industrial Technology Research Institute (ITRI) and National Chung-Shan Institute of Science & Technology (NCSIST).

DeepMentor's mastery in deploying AI models to multiple heterogeneous hardware platforms, such as FPGA, GPU, ASIC, MCU, SoC, and more, showcases their technical expertise in cross-domain integration. This flexibility ensures DeepMentor's solutions can achieve optimal performance and resource utilization across diverse edge environments, adapting to various client needs and existing infrastructures.

As one of the winners of the AI Core Technologies and Chips category in the IC Taiwan Grand Challenge Batch 3, DeepMentor will participate in the Taiwan Innotech Expo (TIE) 2025, which will be held from October 16 - 18, 2025, in the Taipei World Trade Center (TWTC) Hall 1. DeepMentor will showcase their breakthrough innovations in real-time speech-to-speech multimodal LLM AI IP, and visitors are encouraged to join and see how their innovations will fundamentally transform human-machine interfaces, accessibility, and cross-device interactions.

Register for the IC Taiwan Grand Challenge Batch 4

With the continued success of the IC Taiwan Grand Challenge (ICTGC), the organizers will promote it as a long-term event to support semiconductor innovations from Taiwan and abroad. Registration for the next batch of ICTGC will begin in November 2025, and teams with a focus on semiconductor technologies are encouraged to join and compete.

For more details, please visit:
>>> https://ictaiwanchallenge.org/