Media OutReach Newswire – 3 July 2026 – Aurora Mobile, in collaboration with the Hong Kong Productivity Council (HKPC), recently hosted a hands-on workshop dedicated to the deployment of AI Agents, which successfully concluded on June 23. The event brought together decision-makers and digital transformation leaders from government agencies, public service organizations, retail, and NGOs to tackle a pivotal question:
How can AI Agents evolve from simply answering questions to actually executing business tasks?
Jointly organized by HKPC and Aurora Mobile’s flagship platform GPTBots.ai, the workshop was themed “AI Agents Driving Hong Kong’s Public and Enterprise Services: Deployment Practices, Key Technologies, and Real-World Cases.”
Hong Kong Accelerates AI Development
Tommy Lui, Chief Business Officer of GPTBots.ai, shared insights into Hong Kong’s latest AI policy roadmap, including:
-
HK$100 Million AI Efficiency Fund (AIEEI): Allocated in the 2026-2027 Budget, to be distributed over three years to drive AI adoption across government and public institutions.
-
100 Administrative Processes to Be AI-Enabled by 2026: A concrete target documented in Legislative Council papers.
-
HK$3 Billion AI Subsidy Program: Focused on computing infrastructure and research capabilities.
-
HK$50 Million Public AI Training Initiative: Jointly executed by Cyberport, Science Park, and HKPC, targeting 200+ events and reaching 50,000 participants.
-
HK$1 Billion AI Research Institute: Expected to commence operations in the second half of 2026.
(Sources: 2026-2027 Budget, Legislative Council documents, and official announcements)
While the policy direction is clear, Tommy also highlighted a key challenge for enterprises: many organizations are still stuck at the “Q&A tool” stage—limited to answering FAQs and unable to integrate with business workflows or execute real tasks. Bridging the gap from “conversation” to “execution” requires more than just powerful models; it demands a robust, enterprise-grade technology foundation.
Security and Governance: Addressing Enterprise Concerns
As AI begins to access internal data, invoke system tools, and drive approval workflows, security and data governance become top priorities for management.
Rim Wang, AI Agent Engineering Lead at GPTBots.ai, outlined the enterprise-grade AI security framework, covering content moderation, data anonymization, role-based access control, human-in-the-loop mechanisms, and audit logging. These safeguards aren’t afterthoughts—they must be built into every stage, from data input to output.
“Without robust security controls, enterprises are naturally hesitant to deploy AI,” Rim explained. GPTBots.ai embeds a comprehensive security and governance framework at its core, transforming AI from a ‘black box’ into a highly controllable, auditable business tool.
Deployment Methodology: Diagnosis, Targeted Solutions, Continuous Validation
Jacky Li, another AI Agent Engineering Lead at GPTBots.ai, introduced the “Three-Stage Mapping Methodology”:
-
Process Diagnosis: Identify high-frequency, rule-based, and value-quantifiable processes (such as customer service ticket classification, order review, data integration) as AI entry points.
-
Capability Matching: Configure tool-based, knowledge-based, or multi-modal Agents according to task complexity.
-
Validation and Iteration: Pilot with a single pain point, scale gradually after successful end-to-end execution, and evolve into a scalable “digital workforce” platform.
Jacky has validated this approach across manufacturing, cross-border commerce, and service sectors. He emphasized that the greatest challenge in AI deployment is not technology selection, but precisely identifying which business processes deserve AI automation.
Real-World Cases: Cross-Industry Implementation Logic
The most compelling segment featured real-world scenarios from four distinct industries:
-
Financial Services — Automated Complex Customer Service Tickets:
A major bank deployed an AI Agent that accurately identifies customer intent and automatically invokes core banking systems, completing the full cycle of “intent recognition → data retrieval → rule matching → execution → feedback.” Human intervention is triggered only for exceptions, freeing frontline staff from repetitive queries and significantly reducing customer wait times. -
Real Estate — Omni-Channel Customer Data Integration:
A large real estate group built a unified customer data platform integrating WhatsApp, App Push, Email, and SMS. When a customer inquires about management fees via WhatsApp, the AI not only provides the amount but also proactively shares updates like “Your air conditioner repair from last month—has it been completed?” Service evolved from “answering on demand” to “proactive care.” -
Retail — Automated After-Sales Service Tickets:
A retail brand implemented an AI Agent that automatically generates service tickets, classifies issues, and triggers corresponding workflows. Result: after-sales processing time reduced by over 60%, with staff focusing exclusively on genuine exceptions. -
Social Services — Precision Resource Distribution:
A social service organization uses an AI Agent to identify requests, assess priority levels, and leverage omni-channel delivery to send notifications via the most appropriate channel—WhatsApp for social workers, SMS for service recipients, email for management—while automatically tracking progress. The organization transformed from “passive request handling” to “proactive precision delivery.”
The logic across all four scenarios is consistent: AI Agents handle “understanding and decision-making,” while omni-channel platforms manage “reach and execution.” This is Aurora Mobile’s core value proposition—connecting customer engagement, business processes, and AI execution to help enterprises transition from customer connection to AI-driven business action. EngageLab, as an AI-native customer engagement platform, provides unified customer data, reliable omni-channel delivery, and customer verification capabilities. GPTBots.ai, as an enterprise-grade AI Agent platform, delivers end-to-end AI capabilities from Agent building and knowledge management to security governance. Together, they complete the full cycle from “need identification” to “service delivery.”
Policy Support: Enabling SME Digital Transformation
Representatives from the Hong Kong Productivity Council also introduced government subsidy programs for enterprises. The “SME ReachOut” team helps small and medium-sized enterprises identify suitable government funding schemes, answer application questions, and provide free one-on-one consultations and form review services, encouraging SMEs to leverage government support for digital upgrades.
For details, visit:
https://smereachout.hkpc.org/
For SMEs, the technology roadmap is becoming clearer, and policy support is falling into place. The critical question is: Who can fastest transform AI from “experimental” to “truly operational”?
Aurora Mobile: Defining Pain Points, Delivering Solutions
Aurora Mobile (NASDAQ: JG) is a platform technology company connecting customer engagement, business processes, and AI execution, dedicated to helping enterprises transition from customer connection to AI-driven business action. The company operates two core product lines: EngageLab, an AI-native customer engagement platform, and GPTBots.ai, an enterprise-grade AI Agent building platform.
EngageLab centers on AI Agents, unified customer data, and reliable omni-channel delivery to strengthen customer relationships. GPTBots.ai provides end-to-end AI capabilities spanning Agent building, knowledge management, workflow orchestration, and security governance. Together, they construct a complete value chain from “understanding needs” to “delivering services.”
In Hong Kong’s concentrated markets of finance, real estate, retail, and public services, Aurora Mobile is closely collaborating with industry clients to translate AI Agents from technical concepts into practical business process applications.
When AI transcends “answering questions” and truly begins “executing tasks,” the era of enterprise-grade AI officially begins.
The issuer is solely responsible for the content of this announcement.
