
Top startups from InnoVEX 2025
COMPUTEX is one of the leading tech shows across the world, and this year saw the 10th anniversary of its startup section, InnoVEX. According to the event, InnoVEX saw its scale grow 12.5% over last year, with 450 startups from 24 countries exhibiting at this year’s event.
Now that the dust has settled on the exhibition halls, and the jetlag has subsided, Startups Magazine has complied its list of the most exciting startups that were in attendance at InnoVEX 2025.
AeiROBOT
AeiROBOT is a South Korean robotics startup that started as a research-based spin-out from Hanyang University. It focuses on developing humanoid robots to replace human labour.
ALICE, its flagship humanoid robot that is currently in its fourth iteration, incorporates advanced features such as omnidirectional walking, enhanced vision recognition using deep learning models like YOLOv7, and improved localisation algorithms. These enhancements aim to optimise movement, improve predictability, and maintain balance and stability in dynamic environments.
As well as ALICE, AeiROBOT has other robots in its product line, including AIMY, and EDIE. AIMY is a robot guide that moves autonomously around indoor spaces, providing information to customers. EDIE is a pet-like companion robot that has been designed to form a deep emotional bond with its users to provide emotional stability and comfort.
DeepRad.AI
DeepRad.AI is a Taiwanese startup, founded by Professor Cheng-Yu Chen of Taipei Medical University, that is transforming the future of radiology though AI-driven medical imaging solutions aimed at enhancing early disease detection and preventive care. The startup has launched two core technologies: DeepLung-CAC and DeepBrain-Cognito.
DeepLung-CAC is an AI-powered screening platform that utilises a single low-dose computed tomography (LDCT) scan to detect pulmonary nodules and assess coronary artery calcification risk. The system can analyse scans in under one minute, significantly reducing interpretation time.
DeepBrain-Cognito is a computational model that has been designed for early detection of cognitive decline. It provides personalised assessments for populations with suboptimal health, such as the elderly or those with chronic conditions.
EndoSemio
EndoSemio is a Taiwanese medical device startup specialising in the development of miniaturised, single-use endoscopic imaging technologies. Its designs are aimed at enhancing surgical precision, patient safety, and healthcare efficiency.
Its innovations focus on integrating high-definition CMOS imaging sensors with ultra-fine, medical-grade tubing, to enable the creation of compact, disposable endoscopes. The single-use design eliminates the need for reprocessing, thereby reducing infection risks and improving operational efficiency. EndoSemio aims to address the challenges associated with reusable endoscopes, including the labour-intensive cleaning processes and potential for cross-contamination.
NSS Water
NSS Water is a Swedish startup, founded in 2020, that specialises in pioneering watertechnology, delivering nanopure water solutions for high-tech industries. It has developed a patented Water Enhancement Tool (WET) that produces Nanopure Water (NPW), the purest water with no particles larger than 5 nanometres that enables advancements in semiconductor manufacturing. This advanced point-of-use system enables industries like semiconductor manufacturing to enhance production efficiency by reducing water usage by up to 90%, lowering chemical consumption, and minimising environmental impact.
The currently water used in the semiconductor manufacturing process is ultrapure water, and is used for a variety of purposes, ranging from equipment cooling to wafer surface cleaning. A typical 200mm wafer fab that processes 20,000 wafers a month can use up to three million litres of ultrapure water, which is equivalent of the daily water requirements of 20,000 people. While UPW has been the conventional material, it still has some contamination from <10,000 nm particulates, which can still cause product contamination or impact process efficiency, making even purer water necessary to improve the wafer manufacturing success rate, reduce water usage, and reduce manufacturing cost.
semiQa
semiQa is a Polish semiconductor startup that was founded in January 2025, that focuses on developing Analog Neural Networks (aNNs) for neuromorphic computing, aiming to enhance data efficiency and reduce latency in processing large datasets.
semiQa's core innovation lies in its aNN technology, which leverages analog circuitry to mimic neural network functions. This approach offers potential advantages over traditional digital systems, including lower power consumption, reduced latency, and improved data density.
It operates through three primary channels:
- Intellectual Property (IP) marketplace: offering ready-to-use analog design IPs for integration into third-party systems
- B2B and B2G partnerships: collaborating with businesses and government entities to develop tailored solutions
- Joint research initiatives: engaging in collaborative research projects to advance analog computing technologies
Done for another year
COMPUTEX and InnoVEX 2025 were truly exciting events, with hundreds of innovations on display, and a host of keynotes from industry pioneers.
We can’t wait to see what the 2026 edition of the show will have in store, as it returns from the 2-5th June 2026. Following the success of 2025, the show will expand beyond Nangang Exhibition Halls 1 and 2 to include TWTC Hall 1, creating a dedicated AI-powered tech lifestyle ecosystem unique to COMPUTEX.
For more startup news, check out the other articles on the website, and subscribe to the magazine for free. Listen to The Cereal Entrepreneur podcast for more interviews with entrepreneurs and big-hitters in the startup ecosystem.