Original · Unique Research · 2026-03-12
Editor's note: This is the complete English rendition of Unique Research's historical interview with Yang Hongkai. Dudao Technology and Yang Hongkai are romanizations of the Chinese names; DearLink.Plus retains the source's product spelling. The bank scene, Tsinghua backgrounds, capabilities, order maturity, results-based contracts, financing plans and comparisons with OpenClaw are source or company claims, not independently audited deployment or performance findings. The source supplies no quantified ROI, contract terms or compliance assessment. The opening description of customers not knowing they are speaking to AI is retained as reporting; it does not establish consent, lawful disclosure or regulatory compliance, and this translation does not authorize concealed AI outreach or access to customer data. Projections and event references reflect March 2026 rather than current offers or guarantees.
Unique Awards · Guest Interview
Pay for Results + Share the Risk
This AI Company with Tsinghua Roots Is Turning Sales into an Agent-Driven Process
"
AI will inevitably take over process-based work in sales. ROI determines that.
Late at night on Financial Street, hundreds of conversation windows flicker simultaneously on a customer-service screen at a bank's credit-card center. At the other end, customers do not know that they are already talking to AI.
This is an everyday scene at Dudao Technology. With a core team drawn entirely from Tsinghua, the company is using AI Native sales agents to reshape sales productivity in finance.
Not Replacement, but a Sharper Division of Labor: AI Handles Processes, People Handle Emotion
Yang Hongkai, co-founder & COO of Dudao Technology, makes a direct judgment:
“AI will inevitably take over process-based work in sales. ROI determines that.”
But taking over work does not mean jobs disappear. In Dudao Technology's understanding, the division of responsibilities between AI and human salespeople will become clearer.
AI handles:
· Outbound phone calls and customer outreach and communication through WeCom
· Preparation of sales materials
· Customer identification and segmentation
· Analysis of demand data
· Development of personalized sales strategies
People handle:
· Highly nonstandard needs
· Stages requiring face-to-face sales
· Complex deal-closing processes
Human salespeople will become much more efficient, using AI Agents to carry out more sales work in parallel and concentrating their energy on high-value interactions with a strong emotional component.
DearLink.Plus: The AI Native Sales Agent Within a Product Portfolio
Dudao Technology's core product takes the form of an AI Native sales agent, integrated into a portfolio that includes DearLink.Plus, supplying comprehensive AI sales-productivity solutions to clients in finance, consumer industries, and other sectors.
“Compared with a general-purpose CRM or AI assistant, we are not a traditional record-keeping or support tool. Our differentiation is that our product genuinely carries out sales actions with a purpose, while possessing a deep understanding of the industries we serve.”
Specific capabilities include:
· Real-time intelligent conversations that analyze user intent and emotion
· More precisely targeted follow-up proposals based on user intent
· Industry knowhow
· Autonomous SOP creation
· A closed loop for self-learning
· Integration into customers' existing work systems
The target customers are also clear: licensed financial institutions, including banks and insurers; consumer-goods companies; and BPO businesses serving these clients.
Why Do Customers Not Build Their Own AI Sales Teams?
In the AI Agent era, every industry lacks implementation experience and successful examples of applied Agents.
“In the AI Agent era, many customers feel a sense of urgency, but they actually lack sufficient experience and successful examples to draw on when implementing AI Agents in sales,” Yang Hongkai said. “In particular, as large-model capabilities have matured, it is only in recent months that Agents have truly reached the level of humans in executing work. How to optimize organizational structures and workflows on that basis is even more a matter of feeling our way across the river.”
Building an AI sales team in-house carries excessive risks: implementation may fail, and organizational change is painful. Financial clients, in particular, demand extremely high business stability. A WeCom account suspension, complaints, or compliance issues in communications can all have a major impact on operations.
“When a company like ours, with implementation experience and technical capabilities, can empower these customers, they are highly willing to accept the help.”
The Hardest-to-Replicate Capability: Data Accumulation + Industry Knowhow + Shared Operating Risk
“The core is still the growth flywheel that turns accumulated data into industry knowhow,” Yang Hongkai said.
As large-model capabilities mature, the window of technological leadership becomes shorter. The real commercial barrier is a deep understanding of customer use cases and processes: knowing how to apply AI Agents to serve sales in those settings, how to improve human-AI collaboration, and how to use AI to formulate and execute better sales strategies.
But Dudao Technology goes a step further: it shares operating risks with its customers.
“Demonstrating AI's value through actual cost reductions and better results is what will create a barrier for AI Agent companies in the future large-model era. We are therefore willing to share operating risks with customers, rather than simply delivering a system or a trained Agent and stopping there.”
This is not just empty talk. Dudao Technology is advancing a pay-for-results business model and already has established orders in place.
The central indicators of this monetization path's health are the share of revenue from the new business model and growth in the proportion of customers actually paying for results.
“Of course, the market also needs to gradually accept the new payment model. It will be a process, but we believe this is certainly the future trend.”
On OpenClaw: The “Stability Threshold” for Business Applications
Yang Hongkai takes a measured view of Agents such as OpenClaw that run locally and connect to multiple platforms:
“Business applications' requirements for stability and data security make it difficult for OpenClaw to satisfy enterprises' system-level needs. At present, OpenClaw is still centered on improving personal productivity, so it is not a competitor for us. Enterprise applications need more stable, compliant, and predictable results. OpenClaw's explosive popularity does have value in helping the industry understand Agents.”
The Core Development Tasks for 2026: Deeper Industry Expertise and Organizational and Ecosystem Upgrades
His year-end prediction in one sentence:
“I hope that by the end of 2026, Dudao Technology will become a leading provider of AI productivity in China, and a leading company in vertical industries such as finance.”
To achieve that breakthrough, Dudao Technology will prioritize hiring people with combined experience in finance and technology.
At the Unique Awards, Yang Hongkai most hopes to connect with channel agents, enterprise customers, and investors. “We plan to launch our Series B financing this year and look forward to in-depth discussions with investors.”
This article was compiled from Unique Research's in-depth interview with Yang Hongkai.
Selected Interview Q&A
On the Company and Its Products
Q: Introduce yourself in one sentence. Which three indicators do you focus on most in your day-to-day work as COO?
A: I am Yang Hongkai, COO of Dudao Technology. Dudao Technology is a technology company whose core team is composed entirely of people with Tsinghua University backgrounds. We primarily provide AI productivity to industry clients in the artificial-intelligence AI Agent field. As COO, I currently focus most on product-iteration progress, order growth, and customer feedback.
Q: What is the most painful customer problem DearLink.Plus currently solves?
A: DearLink.Plus is actually one part of Dudao Technology's product portfolio. Our present focus is a comprehensive AI sales-productivity solution for financial clients that incorporates DearLink.Plus. We focus on the high labor costs and low ROI that financial companies currently face in reaching, activating, and engaging consumer users.
Q: What differentiates you from a general-purpose CRM or AI assistant?
A: We are not a traditional record-keeping or support tool. Our core differentiation is that our product genuinely executes sales actions with a purpose while possessing a deep understanding of the industries we serve. Specific capabilities include real-time intelligent conversations that analyze user intent and emotion; more precisely targeted follow-up proposals based on that intent; industry knowhow, autonomous SOP creation, and a self-learning loop; and integration into customers' existing work systems.
On the Moat
Q: What is the main reason customers choose Dudao Technology rather than build their own AI sales teams?
A: Customers actually lack sufficient experience and successful examples to draw on when implementing AI in sales. Building their own AI sales teams involves a high risk of failure and high trial-and-error costs. When a company like ours, with implementation experience and technical capabilities, can empower them, they are highly willing to accept the help.
Q: What is the hardest-to-replicate capability in AI sales enablement?
A: The core is still the growth flywheel that turns accumulated data into industry knowhow. As large-model capabilities mature, the window of technological leadership becomes shorter. Deeply understanding customer use cases and processes, knowing how AI can better serve sales in those settings, improving human-AI collaboration, using AI to formulate and execute better sales strategies, and demonstrating AI's value through actual cost reductions and better results—these are the barriers for AI Agent companies in the future large-model era.
On Commercialization
Q: What are your main monetization methods today, and which indicators do you use to judge their health?
A: We are mainly advancing a pay-for-results business model and already have established orders in place. We primarily assess it through the share of revenue generated by the new business model and growth in the proportion of customers actually paying for results.
Q: Which parts of the sales process has AI replaced, and which human roles must be retained?
A: AI mainly replaces traditional manual outreach and communication with customers through outbound phone calls, WeCom, and similar channels, as well as preparation of sales materials, customer identification and segmentation, analysis of demand data, and development of personalized sales strategies. On the human side, the core role today is still resolving highly nonstandard needs and handling sales stages requiring stronger emotional connections.
On AI Collaboration and Human Value
Q: As AI takes over more sales stages, will the core value of an individual salesperson be diluted?
A: AI will inevitably take over process-based work in sales; ROI determines that. But it is not very likely to take over sales that require stronger emotional connections, such as face-to-face selling and complex deal-closing stages, in the short term. We believe human salespeople will become much more efficient, using AI Agents to conduct more sales work in parallel and focusing their energy on high-value interactions with a strong emotional component.
Q: Why focus primarily on financial clients?
A: Financial clients such as banks and insurers have extensive sales needs involving consumers, particularly long-tail consumer customers whom they previously lacked the personnel and energy to serve. They also urgently need better sales results. Through serving them, we have identified common features in these needs, which align very well with our accumulated technology and industry experience.
On OpenClaw and the Agent Ecosystem
Q: How do you view the impact of locally running, multiplatform Agents such as OpenClaw on enterprise AI sales tools?
A: Business applications' requirements for stability and data security make it difficult for OpenClaw to satisfy enterprises' system-level needs. At present, OpenClaw is still centered on improving personal productivity, so it is not a competitor for us. Enterprise applications need more stable, compliant, and predictable results. OpenClaw's explosive popularity has considerable value in advancing the industry's understanding of Agents.
Q: Which stage of sales will become Agent-driven first?
A: At present, we can expect consumer-facing sales stages to become Agent-driven earlier.
On 2026
Q: Where do you see the most promising incremental opportunities in the next 12 months?
A: We are most optimistic about applying AI to transform traditional sales use cases—for example, the many stages of sales work still performed through “people + traditional CRM.” We are very willing to partner for mutual benefit with BPO companies that already have customer relationships.
Q: In one sentence, what do you predict for Dudao Technology at the end of 2026?
A: I hope that by the end of 2026, Dudao Technology will become a leading provider of AI productivity in China, and a leading company in vertical industries such as finance.
Materials drawn from guest interviews for the Unique Awards.