Key Takeaways:
Deepexi Tech is moving to productize bespoke enterprise AI services, a strategy that could lower costs for customers but positions it against the deep engineering model validated by US giant Palantir. The company’s new “Ontology Large Model” aims to solve implementation challenges for large firms, though investors reacted cautiously, with the stock falling over 6 percent.
The core solution to difficulties in enterprise AI implementation lies in the "ontology large model," which transforms high-cost, manually customized projects into standardized products, founder and CEO Zhao Jiehui said in an article published in mainland media. This, he noted, provides a clear pathway for moving from "concept demonstration" to “creating tangible results.”
The announcement comes as shares of Deepexi Tech (01384.HK) fell 6.061 percent. The new model is built on a dataset accumulated over eight years from serving nearly 400 large clients and includes 108 "business ontologies" for industries like manufacturing, retail, healthcare, and finance. The company claims this proprietary data creates a moat that pure model developers cannot easily replicate.
The strategic pivot is designed to shift Deepexi from a service-intensive business to a more scalable, subscription-based model. While this could enhance long-term profitability and create a recurring revenue stream, the immediate stock decline suggests investor apprehension about execution risk or the company's current performance in a competitive market.
According to Zhao, the company’s previous approach to enterprise AI applications was similar to Palantir’s, relying on frontline deployed engineers to perform manual customization at client sites. This deep engineering service model, while effective, is costly and difficult to scale.
Through the new ontology large model, Deepexi Tech has successfully distilled and "modelized" these labor-intensive tasks. This radically reduces customization costs per client, allowing the company to pioneer a product-focused pathway in the Chinese market and transition its business model toward subscription-based recurring revenue.
This article is for informational purposes only and does not constitute investment advice.
