Building VEQRA AI: How I Resolved Enterprise Incidents in 13 Seconds with Qwen3-235B
Building VEQRA AI: How I Resolved Enterprise Incidents in 13 Seconds with Qwen3-235B
构建 VEQRA AI：我如何利用 Qwen3-235B 在 13 秒内解决企业级故障
VEQRA is an existing Microsoft 365 automation platform I built that detects and routes enterprise incidents. But it couldn’t answer the critical questions: Why did this happen? What is the financial impact? What should we do right now? The Qwen Cloud Global AI Hackathon was the opportunity to build the intelligence layer VEQRA was missing.
VEQRA 是我之前构建的一个 Microsoft 365 自动化平台，用于检测和路由企业级故障。但它无法回答一些关键问题：为什么会发生这种情况？财务影响是什么？我们现在应该做什么？Qwen Cloud 全球 AI 黑客松为我提供了一个机会，去构建 VEQRA 所缺失的智能层。
What it does: VEQRA AI orchestrates three specialized AI agents that resolve a critical enterprise incident in 13 seconds:
- Memory Agent — searches historical incident database, identifies similar past cases, determines root cause with confidence score
- BI Agent — calculates financial impact, projects SLA breach risk, assigns criticality score
- Action Agent — generates structured action plan: Teams task, email to Data Owner, Power BI dashboard update
它的功能：VEQRA AI 协同三个专门的 AI 智能体，在 13 秒内解决关键的企业故障：
- 记忆智能体 (Memory Agent) — 搜索历史故障数据库，识别类似的过往案例，并以置信度评分确定根本原因。
- 商业智能智能体 (BI Agent) — 计算财务影响，预测 SLA 违约风险，并分配严重性评分。
- 行动智能体 (Action Agent) — 生成结构化的行动计划：包括 Teams 任务、发送给数据所有者的电子邮件以及 Power BI 仪表板更新。
Demo scenario: a critical Leasing VIP contract (€120,000 OVERDUE) is resolved in 13 seconds with zero human intervention.
演示场景：一份关键的租赁 VIP 合同（逾期金额 12 万欧元）在零人工干预的情况下，于 13 秒内得到解决。
How I built it: Each agent calls Qwen3-235B directly via Alibaba Cloud DashScope, using the OpenAI-compatible endpoint. Agents communicate through structured JSON outputs, coordinated sequentially by a Python orchestrator.
构建方式：每个智能体都通过阿里云 DashScope 直接调用 Qwen3-235B，并使用兼容 OpenAI 的接口。智能体之间通过结构化的 JSON 输出进行通信，并由一个 Python 编排器按顺序协调。
Challenges: Designing prompts that produce consistent, structured JSON outputs across all three agents was the hardest part — keeping the total resolution time under 15 seconds required careful prompt engineering.
挑战：设计能够让所有三个智能体都生成一致且结构化 JSON 输出的提示词（Prompt）是最困难的部分——为了将总解决时间控制在 15 秒以内，需要进行精细的提示词工程。
What I learned: Depth beats breadth. One perfect scenario executed flawlessly is more valuable than five agents that partially work. Qwen3-235B’s long-context reasoning and structured output capabilities made the multi-agent orchestration reliable and deterministic.
心得体会：深度胜过广度。一个完美执行的场景比五个只能部分工作的智能体更有价值。Qwen3-235B 的长上下文推理能力和结构化输出能力，使得多智能体编排变得可靠且具有确定性。
GitHub: https://github.com/nabilfattouch1/VEQRA-AI Demo: https://youtu.be/zakd6bsdzDA