From AI+BI to Enterprise AI Decision Intelligence: Introducing DecideX

AI has been a hot topic in enterprises for over a year, yet few implementations reach core decision‑making. The bottleneck is not model strength—large models can now generate reports and answer most business questions—but the lack of AI that can deliver accountable conclusions and take responsibility for results.

“In the past decade, enterprise data analysis aimed at seeing data; in the next decade, data will be the starting point for decision intelligence.” – Su Chunyuan, Founder & CEO of GuanYuan Data

DecideX is positioned as a native decision‑intelligence platform for enterprise operating scenarios, differentiating itself from generic agents, bots, and knowledge‑base Q&A tools. Its core task is not merely answering questions but supporting decision execution .

Architecture Overview

The platform is organized into three layers:

Top Layer – General Decision Entry: Includes Agent BI, Chat BI, and Insight agents, providing visual interaction, natural‑language Q&A, and proactive insight. These complement, rather than replace, traditional BI.

Middle Layer – Decision Core (DecideX): Acts as the "brain" that transforms existing BI assets, analytical processes, and business experience into callable Agent decision contexts, enabling complex operational decisions.

Bottom Layer – AI‑Ready Enterprise Data Foundation: A trustworthy data base that supports AI decision‑making, with strong data‑engineering, governance, and operational capabilities, and a semantic asset layer for reusable, open‑callable metrics.

What Is “Decision Context”?

Decision context captures the "business truth": metric definitions, role‑specific goals, external signals, historical handling of similar cases, and validated strategies. This information often resides outside databases—in meeting minutes, documents, chat logs, and frontline experience. By making this implicit context explicit and structured, the platform creates callable Agent representations of the enterprise’s "battle map".

Real‑World Cases

At the launch, Unilever China’s Digital Operations Director shared a supply‑chain scenario where traditional experience‑driven operations face order, SLA, capacity, and inventory constraints. Using DecideX, Unilever achieved:

+50% increase in inventory handling capacity

24% reduction in overall logistics cost

~2× improvement in fulfillment efficiency ("order‑to‑ship" speed)

Other customers such as LaiYiFen and QiuTianManMan highlighted that the real challenge is not AI technology but organizational consensus and responsible AI that can be embedded into workflows.

5A Path to Decision Intelligence

GuanYuan proposes a pragmatic methodology rather than a data‑first approach:

Agile: Build a prototype in one day, launch a trial in a week.

Applied: Integrate into real business workflows within a month.

Automated: Enable AI to proactively detect problems and diagnose causes.

Actionable: Convert insights into concrete actions that drive decisions.

Adaptive: Use each action’s outcome to continuously improve the next decision.

This methodology emphasizes scenario‑first, high‑frequency, large‑scale, and measurable AI use cases, arguing that waiting for perfect data is a mistake.

Market Context

Companies such as Palantir, Dataiku, and AWS are also extending toward decision‑intelligence platforms, and Gartner’s 2026 technology‑maturity curve already places “decision‑intelligence platforms” as a mature category.

GuanYuan’s nine‑year history serving over 1,000 customers across retail, manufacturing, and finance, plus early collaborations with Fortune‑500 firms on machine‑learning projects, gives it deep industry know‑how that goes beyond generic AI models.

In summary, the next competitive wave in enterprise software will shift from visualizing data to turning data into accountable actions. DecideX aims to be the bridge that brings enterprise‑specific decision context into AI, creating a closed loop of insight → action → verification.