Haidian's “AI Employees” Open for Registration: Enterprise Agents Move from Pilot Demos to Widespread Adoption
SOURCE: Haidian News
TIME: 2026.11.04
Drafting weekly reports, sorting datasets, conducting operational analysis, identifying equipment anomalies... your AI colleague is here. Recently, DeepWorks, an enterprise agent platform developed by Haidian-based firm Deepexi Technology, officially opened self-registration to all users. No invitation code is required. Users may download the client, register an account and gain access to an “AI colleague”.
Market observers note that this full launch marks a milestone for Deepexi Technology. Beyond custom enterprise solution delivery, the company has formally established a scalable service model centered on standardized cloud-based platform services, taking its AI products into full-scale commercial operation.
“AI Employees”
From Custom Deployment to Standardized Services
What is an enterprise agent? Deepexi Technology defines it as an “AI employee”: an agent that can understand corporate business logic, autonomously plan tasks, invoke tools and deliver final outputs.
DeepWorks is a client-cloud collaborative, local-first AI employee platform built for individual and enterprise users. It converts corporate knowledge into two knowledge layers: a knowledge base Wiki and business ontology. Integrated with existing enterprise systems and empowered by large language models and tooling capabilities, the platform supports scenarios including data analytics, manufacturing, office workflows and operational decision-making. Its four core strengths are AI-native workspaces, an enterprise knowledge hub, multi-role AI colleagues and an open skill ecosystem. Multi-agent collaboration is supported, with integrations to various business tools via SkillHub, MCP and other extension frameworks.
Vertical specialized models have already been deployed on the platform, including DeepSense for manufacturing expertise, DataSense for commercial operations and FDESense for data governance. These cover use cases such as operational analysis, merchandise planning, equipment diagnostics and remote maintenance, enabling coordinated work among multiple specialized roles. Users submit tasks in natural language to trigger the full workflow from requirement intake to result delivery.
As it expands its functional capabilities, the platform has prioritized data security and deployment architecture in its design. Built on a client-cloud collaborative, local-first framework, DeepWorks prioritizes on-premises processing of enterprise data, balancing intelligent functionality and data security requirements.
Alongside open registration, DeepWorks has rolled out distinct service pathways for individual and enterprise users. Registered users are automatically eligible for the personal edition with a complimentary quota of free usage credits. Enterprise users may enquire about enterprise-grade packages to build AI employee platforms tailored to their business needs. Once free credits are exhausted, subsequent usage is billed against purchased tokens. Industry analysts comment that this model creates revenue streams directly tied to ongoing platform consumption, with potential for simultaneous growth in user base and commercial value.
Eight Years of Evolution
From Data Governance to Enterprise Agents
The public launch of DeepWorks represents the latest stage in Deepexi Technology's eight-year development journey. The firm's strategic focus has gradually shifted from data governance toward enterprise agents. In its early days, Deepexi specialized in data governance. Leveraging its self-developed lakehouse architecture, the company concentrated on core workflows for data storage, integration, governance, development and analytics, launching FastData, an enterprise data intelligence suite offering full-stack capabilities from data warehousing to insight generation.
“As our client engagements deepened, we recognized that standalone data governance could no longer meet enterprises' urgent demand to capture tangible AI value,” a Deepexi representative told reporters. To address critical pain points in enterprise AI adoption, the company built two foundational infrastructure layers: FastDataFoil, a data fusion platform, and Deepexi, an enterprise large model platform. FastDataFoil manages multi-source structured and unstructured data and performs tokenization on complex corporate datasets. Deepexi fine-tunes base models through supervised fine-tuning (SFT) and reinforcement learning (RL), enabling deep alignment with private enterprise data to build proprietary models equipped with domain expertise and precise reasoning. Built on this foundation, FastData and FastAGI focus respectively on high-quality knowledge base construction and business process automation, delivering end-to-end support for enterprise LLM deployment.
Following its listing in October 2025, the company accelerated the platformization and scaling of its AI offerings. Over the past year, it upgraded its product stack from FastAGI to DeepexiOS, an AI-native enterprise operating system anchored by enterprise large models, semantic foundations and the enterprise agent platform. This evolution shifts corporate AI from isolated point tools toward governable, continuously upgradable organizational productivity. The full opening of DeepWorks is a concrete delivery of this vision.
Full Launch
Competing on a Global Level Playing Field
DeepWorks' general availability coincides with recent moves by overseas technology vendors. Earlier in October, Google Cloud released Gemini Agent, a general-purpose enterprise agent similarly designed to autonomously decompose tasks and execute sequential tool calls. This release validates broad industry consensus around the enterprise agent track. One key distinction is that Gemini Agent remains in preview, with commercial release timelines and pricing unannounced. DeepWorks, by contrast, is now open to all users for real-world validation. Its product maturity and commercial roadmap are closely aligned with the practical needs of domestic enterprises and individual users.
Capital markets responded positively to the announcement. On October 21, Deepexi's share price surged by nearly 20% at one point during intraday trading.
“By opening DeepWorks broadly, we are translating years of data technology expertise and industry best practices accumulated serving leading enterprises into agent services, delivered as a platform for individuals and companies alike,” the representative stated.
Industry observers regard the shift from project-based delivery to platform services as a pivotal transformation of Deepexi Technology's business model. As corporate AI adoption keeps rising, this “AI colleague” born in Haidian is poised to reach countless ordinary businesses, bringing smart office workflows and intelligent manufacturing out of showcase pilot environments and into widespread real-world use.
返回上一页