Jiushi Intelligent: Making "Chinese AI Solutions" the Global User Operation Infrastructure | AI Competitiveness Interview

[Ebrun Original] When a certain cosmetics brand was at a loss during the goddess festival due to high error rates in manual customer service, AI took over 80,000 pre-sales order reminders within 3 hours, increasing the conversion rate by 40%; when a domestically produced fantasy mobile game faced user loss after a three-month overseas beta test, AI helped achieve a 10.2% recall rate, with a user recall cost of less than 4 yuan, an 11-fold increase in overall ROI for the recall event… These seemingly "magical" scenes are true examples of Chinese AI companies empowering the global development of brands.

Founded in 2018, Guangzhou Jiushi Intelligent Technology Co., Ltd. (hereinafter referred to as Jiushi Intelligent) is one of the promoters behind these "magical" scenes. With self-developed NLP (Natural Language Processing), ASR (Automatic Speech Recognition), TTS (Text-to-speech) and large-scale model technologies, its core product, the "Large-scale Model Contact Center," is redefining user operation rules.

Its intelligent SaaS platform acts as an "AI command center": in the financial sector, it dynamically adjusts debt collection strategies through emotion recognition, increasing the commitment repayment rate to 33%; in e-commerce scenarios, automatic logistics notifications can reduce 75% of after-sales consultation pressures; in cross-border trade scenarios, it can use over 20 languages and dialects to overcome time zone and cultural differences, pushing Southeast Asian COD (Cash on Delivery) marketing connection rates up to 80%.

In the face of overseas enterprises, Jiushi Intelligent has observed: AI technology has deeply penetrated the core chain, but data compliance, cultural adaptation, and ethical supervision remain key challenges. Its founder, Liu Siping, judged that in the next few years, the industry will move towards full-process automation, vertical small model explosion, and deepening human-machine collaboration—AI Agents will take over 80% of repetitive work, while "localized small models + compliance frameworks" will become the winning strategy for globalizing brands.

Around the theme "What kind of AI Competitiveness do Global New Brands Need," Ebrun Power invited a large number of cross-border service companies for a series of interviews, sharing how they apply AI technology to improve the level of service for overseas enterprises and help them enhance comprehensive competitiveness. The following is the conversation between Ebrun Power and Liu Siping, founder of Jiushi Intelligent, edited and organized.

Ebrun Power: Please introduce your company and main business.

Liu Siping: Jiushi Intelligent was established in 2018 and is a member of the China Communication Enterprise Association, Guangdong Artificial Intelligence Industry Association, Shenzhen Artificial Intelligence Industry Association, and the Artificial Intelligence Industry Technology Innovation Strategic Alliance. We are an AI company with core technologies including NLP, ASR, TTS, large-scale models, and big data. The developed product, the Large-scale Model Contact Center, is focused on providing intelligent user operation SaaS platforms for enterprise customers by relying on AI calls, large-scale model calls, intelligent messages, and flash messages, aiming to achieve intelligent operation throughout the full user life cycle, and empower clients' business performance growth.

Ebrun Power: What are your products and their main application scenarios? How do you help global brands/overseas enterprises enhance their competitiveness?

Liu Siping: Jiushi Intelligent is dedicated to building a B2C ecosystem, integrating intelligent touch capabilities, industry models, and operational strategies; it has mature cases in marketing customer expansion, user activity, business services, and care scenes across various industries such as finance, e-commerce, and the internet, assisting enterprises in cost reduction and increased efficiency.

Our core application scenarios for the product include:

1. Marketing Customer Expansion: Achieving high-intent customer screening through user profiling segmentation and personalized script design. For example, pushing "exclusive housekeeping services" for high-net-worth customers and emphasizing "limited-time discounts" for price-sensitive customers.

2. Scene-Based Marketing: Combining holiday marketing, new product release, and other scenes to reach users automatically and push customized information.

3. Private Domain Traffic Introduction: Guiding users to add and follow through AI calls within 24 hours of receiving goods to achieve private flow precipitation. For example, a maternal and child brand increased WeChat friend adding rates to 24% through AI calls + RPA automatic powder adding.

4. Silent User Activation: Automatically identifying the level of churn risk through AI, executing differential recall strategies.

5. Intelligent Debt Collection: In the financial sector, dynamically adjust debt collection strategies through emotion recognition, achieving a 33% commitment repayment rate and 5% recall efficiency.

6. Logistics Notifications: In e-commerce scenarios, the automatic notification of logistics information reduces 75% of the after-sales consultation pressure.

7. Birthday Gift Marketing: In the beauty industry, automatically reach member users and guide them to use birthday benefits, achieving a 40% coupon redemption rate.

8. Satisfaction Survey: After service completion, collecting user feedback through AI calls to optimize the service experience.

We help overseas enterprises in the following aspects to enhance competitiveness:

1. Cost Reduction and Efficiency Increase: Processing over 1 million marketing tasks per day, reducing costs by 80%. For example, a beauty brand completing 80,000 pre-sales order reminders through the Large-scale Model Contact Center, increasing the pre-sale conversion rate by 40%.

2. Operation Efficiency Improvement: Reduce manual operations and improve response speed through task automation outsourcing.

3. Cross-Time Zone Service: Solve service interruptions caused by time differences. For example, in cross-border trade, it can cover full-chain scenarios, significantly reducing labor costs.

4. Cultural Adaptation: Support 20+ mainstream languages and multiple dialects for customized sound to meet different regional customer needs. For example, in Southeast Asia and the Middle East, through the Cash on Delivery (COD) marketing model, connection rates reach 70%-80%.

Here are two specific examples: A maternal and infant brand, using the Large-scale Model Contact Center, completed 80,000 pre-sales order reminders during the goddess festival promotion within 3 hours, increasing the pre-sale conversion rate by 40%, with manual customer service only needing to handle 10% of high-value customer inquiries. A well-known education brand, through the Large-scale Model Contact Center, completed numerous functions such as automatic connection to human representatives, playback-to-human, and intelligent dialogue-to-human, effectively improving user reach rates and conversion rates, such as a 25% AI call transfer rate and a seat conversion rate as high as 6.2%.

Ebrun Power: What changes/upgrades has the rapid development of large-scale models and generative AI technologies brought to your products?

Liu Siping: In 2023, Jiushi Intelligent has begun to vigorously promote the integration of AI voice technology and large-scale model technology, and has taken the lead in the industry by releasing a new generation of intelligent user operation solutions based on AI voice, helping millions of enterprises reconstruct user operation systems.

The AI voice intelligent system is based on multiple generic large model industries, embedded with industry knowledge dictionaries and telephone call systems, RPA, and other plugins, breaking keyword limitations and accurately identifying customers' implicit needs through context semantic analysis, achieving one-to-one user outreach. For example, in financial services, it classifies responses from customers such as "reconsidering" and "not needed," by combining context and referencing knowledge dictionaries, to infer the customer's true intent.

Moreover, when facing some questions that deviate from the business scenario, an AI voice intelligent entity can generate response scripts through calling for large model reasoning, realizing the requirement of responding to all questions. This mode of reasoning leads to a qualitative leap in the actual user interaction experience.

In addition, in terms of project launch efficiency, the knowledge dictionary construction speed of the AI voice intelligent entity is also faster. An Internet finance service test data of 100,000 words, including terms, rules, and regulations, can be organized by the entity within a day. If organized in the form of traditional robot telephony manual processing, it would take at least a week just to process the data, and the resolution rate would only reach 40%, still requiring continuous notes afterwards.

Ebrun: How were your technological and product advantages accumulated?

Liu Siping: The technological and product advantages of 94 Intelligence are accumulated through continuous technological research and development, deep cultivation in the industry, and customer demand-driven accumulation from multiple aspects.

1. Investment in technological research and development and innovation ability. 94 Intelligence has self-developed core technologies in the fields of natural language processing (NLP), speech recognition (ASR), intelligent speech synthesis (TTS), big data analysis, etc., forming a complete technological closed-loop. These technologies provide core capabilities for products such as intelligent interaction, data analysis, and user profiling. 94 Intelligence keeps pace with the forefront of AI technology, continuously optimizes algorithms and models, improves technological performance, lowers the threshold for large model training, and promotes technological popularization.

2. Deep cultivation in the industry and scenario-based solutions. 94 Intelligence has accumulated rich industry experience in multiple fields such as finance, e-commerce, government and enterprise, education, and insurance, and is able to provide customized solutions to address pain points in different industries. For example, in the financial sector, the company provides services such as repayment reminders, repayment audit, and old customer mining for banks; in the e-commerce field, it helps enterprises achieve private domain traffic diversion, activity promotion, and reminders for major promotions. The company's product matrix, which includes AI calls, large model calls, intelligent short messages, flash messages, digital people, etc., covers the entire lifecycle operational scenarios of users, satisfying all-round needs of enterprises from user acquisition, activation, retention to repurchase.

3. Customer-driven and product iteration. 94 Intelligence has always been customer-demand-oriented, iterates product features quickly through in-depth cooperation with customers, and understands their business pain points. For instance, for the problem of low efficiency and conversion rate of enterprise user reach, the company has launched a digital human video call solution, significantly improving user attention and conversion rate through image intuitiveness and digital human attraction. The company provides abundant scenario-based speech templates and intelligent response templates, making it easy for enterprise customers to use. Meanwhile, 94 Intelligence has a professional operations team to assist customers in speech development and quality inspection to ensure communication effectiveness.

Ebrun: From the perspective of the industry segment you are in, what is the degree and effect of the application of AI technology and AI tools by global brand/export companies? What changes and challenges are involved in this process?

Liu Siping: From the perspective of the intelligent speech industry, global brands and export companies have shown a deep penetration trend in the application of AI technology and AI tools, with a significant improvement in application effects, but accompanied by multidimensional changes and challenges. The following analysis is elaborated from the four aspects of application degree, effect, changes, and challenges:

1. Global brands and export companies generally embed AI technology in core business segments, such as AI calls, intelligent customer service, voice assistants, real-time translation, etc. For example, through integrating natural language processing (NLP) and speech recognition (ASR) technologies to realize multilingual support, optimized semantic understanding, and improved interaction experience. In the e-commerce field, large model contact centers are used for customer inquiries, order tracking, and after-sales service; in the financial field, they are used for risk assessment, intelligent investment consultancy, and compliance review; and in the medical field, they are used for patient communication, medical record management, and remote diagnosis.

2. In terms of application effects, a large e-commerce promotion campaign often requires the deployment of 300-500 temporary customer service representatives, with a daily processing of over 100,000 outbound call tasks, yet the daily reach by a manual seat is less than 300 calls per person, indicating obvious service breakpoints, such as late-night order follow-ups and cross-border user communication. A certain cosmetics brand spokesperson admitted, "In the temporary outsourcing of customer service teams on International Women's Day last year, the error rate was as high as 15%, and user complaints surged, damaging the brand's reputation." When using 94 Intelligence's large model contact center for fully automated task hosting, over 1,000,000 tasks can be processed daily, executing critical actions throughout the promotion, reducing costs by 80%.

3. With the maturity of large model technology, AI voice intelligent entities (Voice Agents) are redefining the commercial value of user interaction scenarios with more natural interaction, more accurate decision-making, and stronger autonomous capabilities.

4. With the rapid development of AI technology, intelligent speech service providers are enjoying technical dividends while facing challenges from multiple aspects. These challenges not only come from the technological aspects but also need to address market competition, user demands, data security and privacy protection, ethics and compliance, and other dimensions.

Ebrun: What is your observation on the application of AI in the field of brand globalization/cross-border e-commerce? What trends do you expect to see in the next few years?

Liu Siping: The application of AI in the field of brand globalization and cross-border e-commerce is seeing profound changes. Its technological penetration has expanded from a single link to the entire process, driving the industry towards intelligent, automated, and refined operations.

1. Intelligent product selection and market insight: AI helps brands precisely select products and predict explosive products by analyzing sales data, consumer behavior, and market trends on global e-commerce platforms. For example, using large model technology to deeply analyze consumer trends within the category, combined with local climate, holidays, and other factors to predict the next hot products, the accuracy far exceeds manual analysis.

2. Multilingual marketing and content generation: AI generates marketing material in multiple languages, breaking through language and cultural barriers. For instance, using voice cloning and virtual idol technology to create digital IPs that are in line with regional aesthetics, and simultaneously generate multilingual content while reducing marketing costs.

3. AI marketing customer service and user experience optimization: AI marketing customer service systems provide precise marketing services, improving efficiency and user satisfaction.

In the next few years, I believe there will be several trends:

1. Automation of workflows: AI agents will cover the entire process of product release, customer service, and marketing promotion, achieving an "objective-result" automated closed-loop.

2. Deepening of man-machine collaboration: AI will undertake more than 80% of repetitive work, while humans will focus on strategic innovation and solving complex problems.

3. Outbreak of small models in vertical fields: AI small models developed for specific industries and cultural scenes will emerge, providing more accurate solutions. Different industries such as medical, education, finance, and others have significantly different voice interaction demands and require customized models.

4. Localized adaptation: Small models will better adapt to regional languages, cultures, and regulations, enhancing the localized effects of brands.

5. Strengthening of compliance: With the increasing global regulatory focus on data privacy and AI ethics, brands need to ensure that AI applications comply with the laws and regulations of various countries.

AI enters the commercial reality, with brands and cross-border e-commerce taking the lead, 2025 Global New Brand AI Competitive Conference, scheduled for June 6th in Hangzhou. The theme of this conference is "Evolution Choice," co-hosted by Ebrun and Jiayu Capital. Follow us: