Mondrian AI (official web site) is an organization that describes itself as having an “Finish-to-Finish AI Platform for the whole AI improvement course of,” together with a consulting service to place an entire answer into service from a buyer’s enterprise request and wishes.
Mondrian’s Yennefer Enterprise AI Platform (hyperlink) seems very full and handles {hardware} useful resource allocation, AI mannequin deployments, and the information pipeline from the place it’s saved by way of processing to the end-user dashboard.
Though AI is the first focus nowadays, the corporate has constructed experience in information science and information visualization since they have a tendency to make use of an identical stack and are in excessive demand by numerous prospects.
Yennefer is designed to deal with the entire improvement workflow and may interface with commonplace instruments akin to GitHub, the big public clouds the place code and information are. Clients can then construct their information processing with the mannequin of their alternative and outline their system configurations by software program. Lastly, that is all deployed and managed in manufacturing.
Such a platform can enormously speed up how briskly an entire answer might be deployed. Most corporations, even big ones, don’t have the engineering assets and know-how for each enterprise unit or key mission to deploy a significant AI mission.
That’s the place the consulting side of Mondrian AI may give it an edge with particular prospects, particularly those that wish to deploy one thing complicated comparatively shortly. Whereas there are an increasing number of platforms that declare to carry out comparable duties, the enterprise is predicted to offer its personal engineering or consultants. Simply discovering or deciding on these can already be difficult.
Clients who wish to run their AI workflows on-premises could be articulately concerned about Yennefer because it was constructed for that. Many startups like beginning their operations on public clouds for comfort, however massive, established corporations are sometimes stricter about the place their information can reside.
Mondrian should additionally select its initiatives rigorously, as its engineering assets aren’t limitless. It has labored with prestigious corporations akin to Samsung, SK Telecom, Seoul College, and the Incheon Airport.
Curiously, the corporate can also be constructing its personal ChatGPT-like system, optimized for the Korean language. The big language fashions (LLMs) we sometimes hear about are educated with English language content material, and it’s not unusual to listen to builders complain in regards to the lack of datasets to coach LLMs in different languages.
Now, AI-based functions use the RAG (Retrieval augmented technology) structure to construct chatbots that may reply particular or proprietary questions. Nonetheless, RAG depends on a high-performing LLM. Mondrian AI thinks its upcoming LLM can do the job, which could possibly be a strong argument for Korean corporations to undertake the platform.
Constructing a platform with such depth is spectacular for a younger firm. There are a number of transferring components that require cross-disciplinary experience along with manufacturing information as nicely.
Filed in . Learn extra about CES, CES 2024, Enterprise, Korea and Startups.