With the increasing integration of artificial intelligence (AI) in businesses, there is a growing demand for tools and platforms that facilitate the creation, testing, and deployment of machine learning (ML) models. This category of platforms, known as machine learning operations (MLOps), is already well-established in the market, with numerous startups and established companies providing solutions. However, one South Korean MLOps platform called VESSL AI is aiming to stand out by focusing on optimizing GPU expenses through a hybrid infrastructure that combines on-premise and cloud environments. To accelerate the development of its infrastructure, the startup has recently raised $12 million in a Series A funding round. The funding will be used to cater to companies looking to develop custom large language models (LLMs) and vertical AI agents.
Although VESSL AI is a relatively new player in the market, it has managed to secure 50 enterprise customers, including major names such as Hyundai, LIG Nex1, TMAP Mobility, Yanolja, Upstage, ScatterLab, and Wrtn.ai. The startup has also established strategic partnerships with Oracle and Google Cloud in the U.S., further enhancing its credibility. According to Jaeman Kuss An, the co-founder and CEO of VESSL AI, the company currently has over 2,000 users.
VESSL AI was founded in 2020 by Jaeman Kuss An, Jihwan Jay Chun (CTO), Intae Ryoo (CPO), and Yongseon Sean Lee (tech lead). The founders bring with them valuable experience from previous roles at Google, the mobile game company PUBG, and various AI startups. An’s motivation to establish VESSL AI came from his personal experience at a medical tech startup, where he faced challenges in developing and utilizing machine learning tools. The team realized that the process could be made more efficient and cost-effective by leveraging a hybrid infrastructure model.
The MLOps platform developed by VESSL AI incorporates a multi-cloud strategy and spot instances to reduce GPU expenses by up to 80%. In addition to cost savings, this approach also addresses common issues such as GPU shortages and streamlines the training, deployment, and operation of AI models, including large-scale LLMs. An explains that VESSL AI’s multi-cloud strategy allows for the utilization of GPUs from various cloud service providers, such as AWS, Google Cloud, and Lambda. The platform automatically selects the most cost-effective and efficient resources, significantly reducing customer costs.
VESSL AI’s platform offers four main features: VESSL Run automates AI model training, VESSL Serve enables real-time deployment, VESSL Pipelines integrates model training and data preprocessing to streamline workflows, and VESSL Cluster optimizes GPU resource usage in a cluster environment. These features make it easier for organizations to develop and deploy ML models effectively.
The success of VESSL AI has caught the attention of several investors, who participated in the Series A funding round, bringing the total funds raised by the company to $16.8 million. The investors include A Ventures, Ubiquoss Investment, Mirae Asset Securities, Sirius Investment, SJ Investment Partners, Wooshin Venture Investment, and Shinhan Venture Investment. VESSL AI currently has 35 employees in its South Korean headquarters and a San Mateo office in the U.S.
In conclusion, VESSL AI’s focus on optimizing GPU expenses using a hybrid infrastructure has positioned it as a unique player in the MLOps market. By leveraging a multi-cloud strategy and spot instances, VESSL AI has managed to reduce GPU expenses significantly, offering cost savings to its customers. The success of the startup is evident from its growing number of enterprise customers and partnerships with industry giants like Oracle and Google Cloud. With the recent Series A funding of $12 million, VESSL AI is poised to further develop its infrastructure and cater to the increasing demand for MLOps solutions. It will be interesting to see how the startup evolves and contributes to the advancement of AI and ML technologies.
Source link
