:::

【2021 Solutions】 InfuseAI focuses on building the PrimeHub MLOps software platform to lower the threshold for enterprises to introduce AI technology

In light of the rapid development of Artificial Intelligence (AI), InfuseAI has created PrimeHub, a one-stop AI deployment platform software, with the aim of lowering the barriers for AI adoption and assisting enterprises in a successful transformation. InfuseAI Inc., founded in 2018 by senior developers from the Taiwanese open-source community, g0v's Chia-liang Kao, and KKTIX founder Liang-Bin Hsueh, specializes in PrimeHub, the AI deployment platform software.

InfuseAI's COO Liang-Bin Hsueh shares insights on MLOps implementation and application

▲InfuseAI's COO Liang-Bin Hsueh shares insights on MLOps implementation and application

InfuseAI's COO, Liang-Bin Hsueh mentioned that AI project development is an iterative process. During the AI product lifecycle from data collection, development, model deployment to monitoring etc., it contains many technical debts that prolong the project cycle, increase costs, and significantly reduce the benefits. Furthermore, according to statistics, data scientists spend 65% of their work hours on tasks other than model development. The realization of AI not only involves development but also includes maintenance and cross-team collaboration issues. All these factors significantly slow down the speed of AI development to match the company's needs.

Liang-Bin Hsueh points out that the issues arising from the inability of AI development speed to keep up with enterprise needs can be specifically illustrated in the following two aspects:

• Speed of AI model development: In the past, it took 1 to 2 years to complete an AI project. However, as enterprises adopt AI, the project and AI model numbers multiply.

• Operational issues after AI model deployment: The lifecycle of an AI model begins after deployment. As data accumulates, AI models can be retrained to enhance performance. However, as the number of models grows, operational issues and computational resource bottlenecks emerge.

Hence, the MLOps platform—PrimeHub, developed by InfuseAI, encompasses the processes from AI model development, training management, to operational deployment and monitoring, offering a one-stop platform service through a smooth automated AI workflow that enables true enterprise AI implementation.

MLOps ecosystem

In other words, the InfuseAI team continuously adds to PrimeHub Apps by integrating third-party application services, actively collaborating with more manufacturers to seamlessly integrate AI models into PrimeHub, and eagerly anticipates cooperation with more businesses focused on AI technology and SI partners to inspire more applications on the MLOps platform and further promote large-scale AI implementation. Since its formation over three years ago, InfuseAI's clients include Taiwan AI Academy, E.SUN Financial Holdings, Sinopac Financial Holdings, National Taiwan University Hospital, and Chi Mei Hospital. Among these, InfuseAI works closely with Taiwan AI Academy to address various academic needs. Teaching assistants at each branch only need to operate simply in the PrimeHub platform, where all management tasks are automatically completed. Students in PrimeHub’s self-service platform establish a uniform pre-configured environment, allowing multiple deep learning calculations simultaneously, isolated by containers without interference. Additionally, assistants can decide on the data to load based on the course progress, automatically loading course files and datasets when students launch the environment.

Yushan Financial Holdings began intensive integration into AI development in 2018, acquiring GPU computing resources. They discovered the need for robust infrastructure to speed up operations amidst numerous individuals and projects concurrently. They sought a management platform to assist with computing resource management and data authorization.

Liang-Bin Hsueh states that the PrimeHub platform aims to help enterprises scale AI development, reducing model deployment time from days to hours, further facilitated by APIs and APPs to automate and optimize workflows. PrimeHub operates on a yearly subscription basis, promising ongoing optimization of the platform environment and the flexibility to offer customized services to different customers. Currently, it offers three solutions: PrimeHub Enterprise Edition, PrimeHub Deploy—a lightweight model deployment management plan, and PrimeHub Community Edition, allowing users to choose according to their needs.

(This article is a curated selection from the "AI Engineering Online Meetup")

「Translated content is generated by ChatGPT and is for reference only. Translation date:2024-05-19」

Recommend Cases

【解決方案】台灣軟體科技實力媲美國際 Golface智慧服務促高球轉型
Taiwan's Software Technology on Par with International Standards: Golface's Intelligent Services Transform Golf

Compared to Japan, where 90 of golf courses operate without caddies and use an automated service model, golf course management in Taiwan still heavily relies on human labor Facing a labor shortage of up to 70, adopting a site and membership management platform to provide intelligent golf services may be a transformation worth considering for golf course operators 'Taiwan's software technology is comparable to international standards and definitely has the capability to compete in the global market,' says Tsung-Che Liao, co-founder and CEO of Golface, established in 2014 with the vision to leverage technology at its core, aiming to create Taiwan's first golf entertainment platform With over 9 years vested in cultivating intelligent golf services, Liao is well-versed in the nuances of golf course services He has considerable domain knowledge and has launched a comprehensive intelligent golf solution The world's first networked smart golf cart hits the road automation of golf courses is no longer just a dream In mid-May, Golface's newly developed ARES Smart Golf Cat, the world's first networked smart golf cart, officially became operational Equipped with a dedicated vehicle computer mainframe, dual network systems, AI-based visual recognition cameras, and high-precision GPS tracking, golf courses can now confidently allow golfers to drive themselves The system enables real-time monitoring of any driving violations, and the presence of digital consumption traces allows for insurance coverage The procedure is as follows golfers book the cart via a reservation platform, receive a QR code, pay through the platform, and unlock the cart with the QR code at the golf course The golf cart can then be driven onto the course The course management platform can monitor and restrict the areas through which the cart can travel, ensuring it does not leave the paths Upon completion, the cart is returned through a tablet in the cart In instances of any infractions, penalties are applied directly through the user's account, and for severe violations, future access to the carts may be prohibited This achieves the goal of 'automation' ARES Smart Golf Cat is the world's first networked smart golf cart, officially in service since May 2022 'As labor costs continue to rise, recruiting and training caddies are becoming common pain points in the market While Taiwan's courses still employ caddies, there's a 70 labor shortage,' Tsung-Che Liao added This smart golf cart tablet, combined with a mobile app, has become the ultimate smart caddy Golface is striving to complete the last piece of the 'automated golf course' puzzle Amassing digital consumption trails for advanced client segmentation services Starting with consumer needs, Golface has sequentially launched services like the golf cart tablet, mobile app, golf reservation platform, instructional videos Golface TV, golf tourism, and smart carts The smart cart has been operational since May 2022, currently featuring four units with plans for mass production in the latter half of 2022 Although the cart currently requires manual operation by golfers, remote operation is anticipated early in 2023, with autonomous driving expected in the third phase Via the cart tablet and management system, staff can understand the status of the course through on-screen visual representations, showing each cart's real-time and relative location, departure times, and duration of service per hole, which aids course managers in monitoring on-course consumption effectively, thus reducing traffic jams and customer complaints 'Previously, we relied on staff's mental imagery now, we can employ imagery to visualize real-time situations on the course This makes it possible for those who don't understand golf to work in this field,' emphasized Tsung-Che Liao While course control has traditionally been handled by experienced professional players, the shortage of skilled professionals makes hiring even more challenging Therefore, replacing manpower with digital tools yields twice the result with half the effort The golf cart tablet has entered the Japanese golf market, installed at Fukuoka Century Golf Club Golface's golf cart tablet has been introduced to 14 domestic courses, and has now officially entered the Japanese market, favored by Fukuoka Century Golf Club, where tablets have been installed in carts providing automatic voice announcements for hitting strategies, distance measurements, and visual charts displaying hitting data During the COVID-19 pandemic, with borders closed, Golface utilized OTA technology to provide software updates and troubleshooting, ensuring uninterrupted services, which was highly appreciated by the Japanese golf courses Tsung-Che Liao remarks that Taiwan's software technology is not inferior to other countries like Japan, but more support from golf courses is needed to help transform the industry intelligently 'To assist in the transformation of golf courses, the first step is digitalization,' Liao pointed out By helping courses accumulate data and understand customer service cycles and hitting rhythms, it enables courses to avoid congestion and serve more customers To date, Golface has accumulated data on over 20,000 teams, 35 million scorecards, and over 10 million records This data helps enhance management performance, segment customer layers, reduce complaints, and plan marketing strategies for off-peak periods Golface co-founder and CEO Tsung-Che Liao has spent 9 years deepening intelligent golf services, aiming to build Taiwan's first golf entertainment platform「Translated content is generated by ChatGPT and is for reference only Translation date:2024-05-19」

【解決方案】聲麥無線推出殺手級5G即時AI語音翻譯 降低5成口譯成本
VM-Fi launches killer 5G real-time AI voice translator that reduces interpretation costs by 50%

With "Voice as a Service VaaS", VM-Fi Inc launched the world's leading 5G real-time AI voice translation service "VM-Fi," providing AI real-time translation subtitles TranSpeech and AI multi-lingual smart counter TransDisplay The service is applied to real-time translation services in diverse fields, such as international exhibitions, tourism industry, and retail stores The translation service can be quickly set up in 15 minutes, which can significantly reduce the customer's interpretation service cost by 50, creating benefits in terms of time and labor costs This killer application is a product launched by VM-Fi, which has been established for three years It can significantly reduce the cost of interpretation by half It successfully entered the Japanese market and is highly popular among consumers TranSpeech real-time subtitles for speeches and TransDisplay smart counter AI service solution For organizers of international forums and exhibitions, paying high interpretation fees has always been a common pain point hard to put into words According to statistics, to hold an all-English forum in Taiwan, including two interpreters, setting up an interpretation booth, on-site radio station, control panel, and audio, the total cost will be at least NT100,000 to meet the on-site interpretation needs Established in September 2020, VM-Fi launched the "VM-Fi 5G real-time AI voice translation service" In just 2 years, and successively won the Mobileheroes of the Ministry of Economic Affairs Industrial Development Bureau, Start-up competition in the Smart City Project, Qualcomm Innovate in Taiwan Challenge QITC, Japan's JR Kyushu Innovative Business Competition Excellence Award, and the CES 2022 Innovation Award It has not only received praises in the market, but also attracted great attention from Taiwanese investors and was adopted by large Japanese trading companies It hopes to help people around the world avoid communication barriers through smart city solutions, and enjoy a convenient life in smart cities 5G real-time AI voice translation does not require expensive interpreters and equipment and reduces the interpretation cost by 50 The VM-Fi team, which has a solid technical foundation in language translation, digital content, and UX development, has learned that only 20 of high-end consumers in the market can afford the high cost of interpretation To meet the remaining 80 of market demand, VM-Fi combines 5G high-speed transmission and AI voice recognition technology help customers reduce labor and costs Its business solutions can be applied in international exhibitions, tourism services, shopping malls, and online and offline business meetings 5G high-speed AI speech translation service process "Unlike traditional interpreters' sentence-by-sentence interpretation, which is a time-consuming and inefficient process that prevents the audience from smoothly listening to a speech," VM-Fi further stated that the real-time AI speech translation service uses AI algorithms to analyze the speakers speech The AI determines the parts of speech and subject of the speaker, and then immediately translates it The lecturer does not need to wait for the interpreter, and only needs to connect the microphone cable to the VM-Fi 5G real-time AI voice translation service to speak freely The audience can immediately read the translated subtitles 5G real-time AI voice translation was used in the Kyoto Smart City Summit amp Expo for two consecutive years in 2022-2023 The "VM-Fi 5G real-time AI voice translation service" currently supports Chinese, English, Japanese, Korean, Spanish, French, and German, providing translation services between these seven languages During the pandemic, when the demand for online activities increased significantly, VM-Fi introduced the world's first real-time subtitle service solution The contents spoken by the foreign speaker is translated in real-time and displayed on the live broadcast screen, allowing the audience to understand contents shared by the speaker in real time The operation method is also very simple Audiences do not need to download the APP All they need to do is open the Youtube or Facebook live broadcast platforms provided by the organizer The real-time subtitle solution not only saves customers the trouble of translation work, but also allows online listeners to enjoy seamless real-time translation services without any worries Real-time subtitles provide audiences with seamless real-time translation service In addition, the smart counter solution TransDisplay launched by VM-Fi in Japan is also very popular among consumers VM-Fi stated that Japan has a large elderly population and a large number of tourists Especially during the pandemic, most consumers wore masks and there was often a gap in communication The smart counter directly converts the communication between two parties from voice to text and displays on the transparent partition, providing the most considerate service that the public can clearly see at a glance In the future, VM-Fi will work with Taiwanese panel manufacturers, integrating software and hardware to provide voice-to-real-time subtitle translation services in Japanrsquos shopping malls, stations, airports, and government agencies Smart counters provided people with friendly communication services during the pandemic User experience is the key to VM-Fi taking root in the Japanese market Facing the explosive growth of cross-border tourism business opportunities after the pandemic, VM-Fi is confident in business expansion VM-Fi stated that the VM-Fi 5G real-time AI voice translation service has demonstrated excellent product and service resilience during the pandemic, and can flexibly meet customer needs for various physical or online services VM-Fi plans to establish a headquarters in Japan before April 2025, and actively cooperates with Kansai and Kyoto prefectural and municipal governments to expand its operations After establishing a solid foundation in the Japanese market, the EU market will be the next important target VM-Fi participated in fundraising activities of the AI Founders Alliance of the AI Program of the Industrial Development Bureau, Ministry of Economic Affairs The company relied on the assistance of the Institute for Information Technology III, Taiwan-Japan Industrial Collaboration Promotion Office TJPO, and the Japan External Trade Organization JETRO to actively seize Japan's market "Japan's market not only values digital transformation, but also user experience UX" Therefore, after the successful launch in Japan, everything will fall into place when it expands to other markets around the world VM-Fi app in Japan's shopping malls, stations, airports, and government agencies VM-Fi is a startup with passion and innovative spirit nbspfounded in 2020 We focus on AI speech recognition and real-time translation technologies, and are committed to allowing people around the world to enjoy high-speed speech translation during international speeches, at service counters, and during communication, completely removing the language barrier and allowing everyone to enjoy equal rights to information VM-Fi's AI service is not only a technological innovation, but also a commitment to global sustainable development A Digital solutions We are committed to reducing paper use Every sheet of A4 paper saved reduces carbon emissions by 78 grams, using practical digital solutions to promote environmental protection B Renewable energy decisions We choose cloud services that use renewable energy and plan to achieve the goal of using cloud services that use 100 renewable energy by 2025, in order to create a greener technology foundation for the future C Water saving decisions We are committed to using cloud services that achieve positive water benefits before 2030, ensuring that the replenishment of water resources exceeds consumption, and contributing to the future of the Earth D Net zero emission decisions When using new data centers, we will choose net-zero deforestation to protect the natural environment E Sustainable development goals Through the decisions above, we are making active efforts to achieve the United Nations Sustainable Development Goals SDGs 4, 7, 9, 10, 11, and 13, exerting every effort to create a better world VM-Fi's vision is to create a world without communication barriers, allowing everyone to communicate freely and jointly move toward a better future Let us work together to contribute to global communication equality and sustainable development, and create a greener and better future nbsp

【解決方案】瑕疵辨識率達百分百 耐銳利科技獲面板大廠青睞
Defect identification rate reaches 100%, Nairi Technology is favored by major panel manufacturers

On the machine tool production line, there are some slight differences in the first step of assembly Accumulated tolerances will cause the assembly work to be repeated, which is time-consuming and labor-intensive, resulting in shipment delays that will impact the company's reputation Narili Technology Company focuses on the field of smart manufacturing and provides various AI solutions It uses machine learning models to inherit the experience of old masters In the CNC processing machine assembly and casting process, it uses AI to analyze production line data, accurately adjust various data, and improve Production accuracy is 25 This AI production line data analysis system is called "Master 40" by Huang Changding, chairman of Naruili Technology It is the most evolved version of the master plus artificial intelligence It has been used in machine tool processing factories with remarkable results In addition, Nairi Technology used AI defect detection technology to participate in the 2021 AI Rookie Selection Competition of the Industrial Bureau of the Ministry of Economic Affairs, assisting AUO in advanced panel image defect detection, with an accuracy rate of 100, and won the award Assisted panel manufacturer AUO to solve problems with 100 accuracy in defect detectionHuang Changding further explained that during the production of general panels, edges and corners are There may be defects in the corners Although the defects are visible to the naked eye, AOI is often difficult to identify, causing the detection error rate to often exceed 30 Therefore, re-inspection must be carried out with manpower to improve the accuracy rate However, in response to the demand for a small number of diverse products and insufficient manpower, using AI detection is indeed a good method Nairui Technology, founded in 2018, has been able to win the favor of major panel manufacturers with its AI technology in just three years In fact, it has been honed in the field of CNC machine tools for a long time Tang Guowei, general manager of Narili Technology, pointed out that the top three CNC machine tool factories in Taiwan hope to introduce AI into the two production lines of assembly and casting Among them, on the assembly line, in order to maintain the accuracy of assembly, every part of the component is designed Tolerances are designed During assembly, each component is within the tolerance However, the cumulative tolerance still fails the final quality inspection and must be dismantled and reassembled This is not only time-consuming and labor-intensive, but also causes waste "After entering the production line, I realized that some masters have accumulated a lot of experience and are good at adjustment After his adjustment, the accuracy rate has improved a lot and the speed is faster" On the contrary, the new engineers did not Based on experience, it takes a long time to adjust and may not pass the quality inspection The yield rate of Master 40 system has increased significantly from 70 to 95Tang Guowei then said that the original size data set by Master during assembly All were recorded on paper After the information was written, it was stored in the warehouse and sealed No one studied the relationship between the dimensions Narili assists customers in designing the Fu 40 system Through the human-machine panel, the master can directly input the measured dimensions and related data during assembly After collecting data from different masters, AI algorithms are used to analyze the relationship between the data and create an AI model The AI model automatically notifies the operator what size to adjust to, and the quality inspection will definitely pass In this way, the yield rate will be improved It has increased significantly from 70 to more than 95 Narili Technology Company focuses on the field of smart manufacturing and provides various AI solutionsTang Guowei added, assembling the spindle of a CNC processing machine It took four hours In the first step, the machine made measurement errors, including vibration, temperature, speed, etc that were out of range It had to be dismantled and reinstalled, which took another four hours How to adjust after disassembly depends on the experience of the master At first, the master may have done the best assembly method based on experience, but the error rate was also 30, and the assembly took several days With the assistance of AI masters, the assembly time only takes half a day, and the yield rate reaches over 95, saving a lot of time and manpower "Use the AI model of machine learning to collect the experience of all the masters and provide it for AI learning The first step is digitalization, and the second step is knowledgeization This is the transformation of the enterprise "An important key", Huang Changding believes that Narili Technology is an important partner in the transformation of traditional manufacturing from automated production to digital transformation In addition, another industry that Naili Technology focuses on is the smart car dispatching system of the leading brand of elevator manufacturers The so-called car dispatch referring to the elevator car means that if there are more than two elevators, group management is required In the past, car dispatching was based on fixed rules If the elevator was closer to the requested car, that elevator would be automatically dispatched On the one hand, it did not take into account that dispatching a car if the elevator was called too many times might make other people wait longer The previous vehicle dispatching model did not take into account the usage characteristics of the building, resulting in a lot of waste For example, in an office building, there are peak hours in the morning, lunch break, and afternoon after work AI smart car dispatch can be flexibly adjusted according to off-peak and peak hours, increasing the efficiency of car dispatch, reducing waiting time, and reducing wasted electricity Introducing elevator smart dispatch to improve transportation efficiency and have environmental protection functionsHuang Changding added that just like the previous traffic lights at intersections, the system has already The number of seconds to stop and pass on highways, sub-trunks and small streets is programmed Smart traffic lights are now used to flexibly adjust waiting times to make road sections prone to congestion smoother Using AI to learn usage scenarios and introducing a smart dispatch system into elevators will improve transportation efficiency and make it more environmentally friendly In addition to introducing smart elevator dispatching, Nairili also introduced AI into the smart production and shipment scheduling system of elevator factories Elevator factories often cannot accurately estimate the customer's elevator delivery date For example, office buildings or stores must be completed to a certain extent before the elevator can be installed on the construction site If affected by unexpected factors such as delays in the customer's construction period, the elevator factory will often be idle or the schedule will be difficult to arrange Tang Guowei pointed out that generally those who understand the progress of client projects may be from business or engineering, but overall, the accuracy rate of shipments is only about 60, which means that 40 of them will not be shipped as scheduled Therefore, if the shipping schedule can be accurately estimated, the production line can be freed up for emergency orders or other product production needs The AI smart scheduling system will analyze past shipment data, about 20-30 parameters such as climate, distance between the factory and the construction site, and customer credit, and put them into the AI algorithm to accurately predict whether shipments can be made as scheduled goods Huang Changding also specifically stated that the machine learning of Naili Technology is not ordinary machine learning, but also incorporates various calculation methods such as traditional image processing technology and statistics Only by being very familiar with the domain knowledge can we make good products AI models are also where the company’s competitiveness lies He emphasized that the data that general SaaS platforms can process is very limited, and the accuracy rate has increased from 70 to 75 at most Naili’s strength lies in AI algorithms and machine learning, and it must be coupled with in-depth industry knowledge to produce output Good AI model Narili Technology started with the AI project, gradually deepened the technology, chose to start with the more difficult tasks, and accumulated rules of thumb It is expected to develop SaaS services this year 2022, based on customer needs starting point, gradually gaining a foothold and becoming an important partner in smart manufacturing The picture left shows the general manager of Naruili Technology Tang Guowei and Chairman Huang Changding right「Translated content is generated by ChatGPT and is for reference only Translation date:2024-05-19」