Zhongke WenGe Unveils AI Decision-Making Product Suite: Reducing Business Decision Time to 10 Minutes

On July 18, 2026, at the World Artificial Intelligence Conference (WAIC), Chinese AI company Zhongke WenGe announced a comprehensive AI decision-making product suite. The system features a five-layer structure—“Foundation, Hub, Core, Brain, and Terminal”—encompassing the entire chain from data governance and business modeling to model inference and agent execution. The company claims this will transform corporate decision-making processes from “heavyweight informatization” to “intelligent assembly.”

Zhongke WenGe CEO Luo Yin emphasized during the announcement, “General large-scale models can only answer questions about the present, but Decitron calculates the future.” This statement reflects the company’s clear vision for AI specialized in decision-making.

Overview of the Five-Layer Structure

The newly unveiled decision-making product suite consists of five layers:

1. Foundation Layer - TokSea Token Platform: This foundational platform, referred to as the “Intelligence Foundation,” enables unified computational capabilities, model operation, intellectual asset measurement, scheduling, and governance.
2. Knowledge Hub Layer - DIP Ontology Data Platform: Positioned as the “Knowledge Hub,” this platform elevates fragmented data into business ontology—akin to “digital twins”—that large models can comprehend. It automates data integration and semantic enrichment.
3. Model Core Layer: This layer comprises two engines—ScienceOne and YaYi. ScienceOne specializes in understanding scientific data, experimental design, and solution optimization, while YaYi focuses on general language comprehension and inference tasks. This dual-engine setup addresses both specialized and general business needs.
4. Decision Brain Layer - Decitron: Positioned as the centerpiece of the announcement, this decision engine specializes in deep analysis, simulation, scenario comparison, and selection.
5. Terminal Scene Layer - Claworks: Responsible for injecting precise intelligent capabilities into on-site operations and processes, serving as the final stage of agent execution.

These five layers are supported by the DOMA (Data/Ontology/Model/Agents) architecture, which progresses systematically from data governance to ontology modeling, model inference, and ultimately agent execution, culminating in actionable decisions.

Performance of the Decitron Decision Engine

One of Decitron’s core achievements is its State-of-the-Art (SOTA) performance across three key indicators in the internationally recognized PolyBench dataset. According to Zhongke WenGe, Decitron outperformed GPT-5.5 and Claude-4.7 in cross-domain stability.

During a 28-day predictive market simulation, Decitron implemented a probabilistic signal-based strategy that achieved an annualized return of 160.28%. While this result was obtained in a simulated environment, it serves as a compelling indicator of the effectiveness of its probabilistic decision-support capabilities.

Far from being a mere predictive model, Decitron is designed to simulate multiple scenarios, compare potential pathways, evaluate risks, and propose optimal solutions. It integrates internal business data with external signals such as policies, markets, industries, and public opinion to support strategic corporate decision-making.

Practical Applications and Benefits

During the announcement, Zhongke WenGe highlighted a case study involving a knit-dyeing company. The system was able to detect equipment anomalies in real-time, identify the root causes by analyzing relationships between orders, machinery, and processes, and predict production progress. It also facilitated equipment efficiency comparisons, maintenance scheduling, notifications, and workflow adjustments. This created a closed-loop process of “sensing → analysis → decision-making → execution → feedback.”

In broader strategic decision-making contexts, the system integrates internal business data with external signals to address challenges such as market entry, production capacity settings, major investments, and strategy adjustments. By automating scenario simulations, pathway comparisons, and risk evaluations, the system claims to reduce the analysis process from several days or weeks to approximately 10 minutes.

According to Zhongke WenGe, their intelligent decision-making system can cut construction and operational costs by over 90% and improve decision-making response times by up to 192 times. Tasks that previously required an average of 32 hours can now be completed in just 10 minutes. However, these figures are specific to certain use cases and may vary depending on the implementation environment.

Applications in Scientific Research

At the industrial R&D level, the ScienceOne engine supports understanding scientific data, designing experiments, and optimizing solutions. Particularly in fields like materials science, the engine aims to shift research and development from traditional trial-and-error methods to a data-, knowledge-, and model-driven approach.

According to Zhongke WenGe, in evaluations involving over 60 specialized scientific research tasks, ScienceOne 2.0 outperformed general-purpose flagship models like Gemini-3.1-pro and GPT-5.5 on most tasks. It achieved international benchmarks in tasks such as chemical property prediction, molecular structure prediction from spectra, and protein site prediction.

Market Position

Data from consulting firm CIC indicates that Zhongke WenGe held an 11.4% market share in the mid-sized enterprise large model-driven decision intelligence market in 2024, ranking first. To date, the company has served over 1,000 companies globally, covering key areas such as data intelligence, operational intelligence, industrial intelligence, and strategic intelligence.

Zeng Dajun, deputy director of the Chinese Academy of Sciences’ Automation Institute, outlined the evolution of decision-making paradigms during an academic report at the event. He identified three stages: the first focused on rational decision-making through mathematical methods, the second on data-driven decision-making, and the current third stage, characterized by artificial intelligence.

Zeng stated, “Previously, many complex decision-making problems were considered unanalyzable, unquantifiable, and uncomputable. Today, thanks to artificial intelligence, these challenges have become solvable, analyzable, and actionable.” He emphasized that decision intelligence has entered a critical phase of system development, where the ability to provide practical tools and frameworks will determine future progress.

Competitive Differentiators

Zhongke WenGe’s product suite takes a distinct approach compared to general-purpose large models. As CEO Luo put it, while models like ChatGPT or GPT-5.5 answer questions about the “present,” Decitron is designed to “calculate the future.” This focus on decision support through probabilistic models and simulation-based reasoning sets it apart.

The end-to-end integration provided by the DOMA architecture—from data governance to decision execution—is another competitive edge. While many AI companies focus on individual tools or models, Zhongke WenGe offers an all-encompassing suite that covers the entire decision-making chain.

However, the decision intelligence market is rapidly expanding, with major tech firms like ByteDance, Alibaba, and Tencent developing similar products. Whether Zhongke WenGe can maintain or grow its 11.4% market share will depend on its ability to innovate and secure widespread adoption.

Editorial Opinion

In the short term, the impact of Zhongke WenGe’s product suite on corporate decision-making processes may be limited but is certainly noteworthy. The claims of over 90% cost reduction and a 192-fold increase in response speed are benchmarks achieved under specific conditions, but broader adoption could invigorate the decision-making AI market. Over the next three to six months, competitors are likely to introduce similar products, leading to greater market competition and consumer choice.

Looking ahead, as decision-making paradigms evolve from “mathematical → data-driven → intelligent,” companies specializing in decision-focused AI, such as Zhongke WenGe, may see their roles expand. While general-purpose large models are being adapted for practical decision-making, the demand for domain-specific decision engines like Decitron will serve as a litmus test for the future of this technology.

Nevertheless, questions around accountability and explainability in AI-driven decisions remain unresolved. The development of regulations and guidelines will need to progress in parallel with technological advancements. A critical question for the editorial team is whether Decitron’s SOTA achievements on PolyBench translate into tangible, high-quality decision-making in real-world business scenarios.

References

• “中科闻歌发布首个AI决策“全家桶”，企业决策卷入“10分钟时代””, by LCC_Beta版 — 钛媒体, 2026-07-18T13:53:45.000Z (ARR)
• Source URL: https://www.tmtpost.com/8070543.html