client-meeting-prep-agent
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseClient Meeting Prep Agent
客户会议准备Agent
Helps relationship managers prepare for client meetings by researching the client, the topic, and internal knowledge to deliver a crisp Point of View (PoV) and comprehensive Speaker Notes.
帮助客户经理通过研究客户、会议主题和内部知识,为客户会议做准备,输出简洁的观点(Point of View, PoV)和全面的演讲备注(Speaker Notes)。
Overview
概述
Purpose
用途
Prepare relationship managers for client meetings by synthesizing research across external and internal sources into:
- Point of View (PoV): A tailored recommendation or solution for the client's specific situation (crisp, actionable)
- Speaker Notes: Detailed supporting information with context, examples, and evidence (5-7 minute read)
通过整合外部和内部来源的研究内容,为客户经理准备客户会议,输出以下内容:
- 观点(Point of View, PoV):针对客户具体情况的定制化建议或解决方案(简洁、可落地)
- 演讲备注(Speaker Notes):包含背景、示例和证据的详细支撑信息(阅读时长5-7分钟)
Key Characteristics
核心特性
- Synthesized narrative: Conflicting sources are synthesized into one coherent perspective, not presented separately
- Client-centric: Tailored to the specific client, person, company, and topic—not generic
- Evidence-backed: Claims in PoV supported by detailed evidence in Speaker Notes
- Multi-source: Integrates web search, LinkedIn social media, and local file system sources
- 整合式叙事:将相互矛盾的信息源整合成一个连贯的视角,而非单独呈现
- 以客户为中心:针对特定客户、对接人、公司和主题定制内容,而非通用模板
- 基于证据:观点中的所有主张都有演讲备注里的详细证据支撑
- 多源整合:整合网页搜索、LinkedIn社交媒体和本地文件系统的信息
Inputs
输入参数
Required Parameters
必填参数
-
Person Name (string)
- Full name of the contact/stakeholder you're meeting with
- Used to search LinkedIn and personal notes folders
-
Company (string)
- Company name where the person works
- Used to search account information and web sources
-
Topic (string)
- The subject matter of the meeting
- Could be: a business challenge, an industry trend, a product discussion, a transformation initiative, etc.
- Used to guide all research and recommendations
-
对接人姓名(字符串)
- 会议对接人/利益相关者的全名
- 用于搜索LinkedIn和个人笔记文件夹
-
公司名称(字符串)
- 对接人所在公司的名称
- 用于搜索客户账户信息和网页来源
-
会议主题(字符串)
- 会议的主题内容
- 可以是:业务挑战、行业趋势、产品讨论、转型计划等
- 用于指导所有研究和建议方向
Optional Parameters
可选参数
- Meeting Date (date): When the meeting occurs (helps with relevance of research)
- Output Length (enum): Target read time—"2-3 minutes", "5-7 minutes", "10+ minutes" (default: "5-7 minutes")
- 会议日期(日期):会议举办时间(有助于提升研究内容的相关性)
- 输出长度(枚举值):目标阅读时长——"2-3分钟"、"5-7分钟"、"10+分钟"(默认值:"5-7分钟")
Outputs
输出内容
A markdown document containing:
一份Markdown文档,包含:
1. Point of View Section
1. 观点章节
- Structure: Clear, compelling narrative (not bullets)
- Content: Tailored recommendation or solution addressing the client's specific situation
- Length: 200-400 words
- Tone: Professional, forward-thinking, confident but not prescriptive
- Evidence: Each claim grounded in research findings (noted via citations within the narrative or as inline references)
- 结构:清晰、有说服力的叙事文本(非项目符号)
- 内容:针对客户具体情况的定制化建议或解决方案
- 长度:200-400字
- 语气:专业、前瞻性、自信但不强加观点
- 证据:每个主张都基于研究结果(通过叙事中的引用或内联参考标注)
2. Speaker Notes Section
2. 演讲备注章节
- Structure: Organized paragraphs with supporting details
- Content: Context, examples, case studies, industry data, competitive insights
- Length: ~800-1200 words (appropriate for 5-7 minute read)
- Subsections (as appropriate to topic):
- Market/Industry Context
- Competitive Landscape
- Client-Specific Insights (based on Company and Person research)
- IBM/Our Unique Value
- FAQs or Anticipated Objections
- Next Steps
- 结构:按逻辑组织的段落,包含支撑细节
- 内容:背景信息、示例、案例研究、行业数据、竞品洞察
- 长度:约800-1200字(适合5-7分钟阅读)
- 子章节(根据主题按需设置):
- 市场/行业背景
- 竞争格局
- 客户专属洞察(基于公司和对接人研究)
- IBM/我方独特价值
- 常见问题或预期异议
- 后续步骤
3. Sources Section (optional but recommended)
3. 来源章节(可选但推荐)
- List of sources consulted (web links, file names, LinkedIn profiles, etc.)
- Helps validate research and allows manager to dig deeper if needed
- 列出参考的所有来源(网页链接、文件名、LinkedIn主页等)
- 有助于验证研究内容,方便经理深入挖掘信息
Information Sources & Search Strategy
信息来源与搜索策略
1. Web Search
1. 网页搜索
- Tool: Web search engine
- Topics: Industry trends, competitive analysis, general PoV on the topic
- Depth: 3-5 searches to build comprehensive understanding
- Queries: Topic + Company context, industry reports, expert perspectives
- 工具:网页搜索引擎
- 主题:行业趋势、竞品分析、关于该主题的通用观点
- 深度:3-5次搜索,构建全面认知
- 查询词:主题+公司背景、行业报告、专家观点
2. LinkedIn / Social Media
2. LinkedIn/社交媒体
- Source: LinkedIn
- Search For:
- The person's profile (background, roles, interests, recommendations)
- Company page (size, industry, recent news, employee engagement)
- Topic-relevant discussions or articles they've engaged with
- Usage: Personalize messaging, identify pain points, understand their perspective
- 来源:LinkedIn
- 搜索内容:
- 对接人的个人主页(背景、职位、兴趣、推荐内容)
- 公司主页(规模、行业、近期新闻、员工互动情况)
- 他们参与过的与主题相关的讨论或文章
- 用途:个性化沟通内容、识别痛点、理解他们的视角
3. Account Box (Local Folder)
3. 客户账户文件夹(本地文件夹)
- Path: (assumed folder structure)
~/account-box/[Company Name]/ - Contents to Search:
- Account summaries, win-loss analyses, relationship notes
- Previous proposals, agreements, communications
- Company org chart, budget/procurement info
- Past meeting notes or feedback
- Usage: Understand relationship history, known pain points, prior conversations
- 路径:(预设文件夹结构)
~/account-box/[Company Name]/ - 搜索内容:
- 客户账户摘要、输赢分析、关系记录
- 过往提案、协议、沟通内容
- 公司组织架构、预算/采购信息
- 过往会议记录或反馈
- 用途:了解合作历史、已知痛点、之前的沟通内容
4. Personal Notes (Local Folder)
4. 个人笔记(本地文件夹)
- Path: (assumed folder structure)
~/personal-notes/ - Contents to Search:
- Documents about the Topic
- Notes about the Person (interactions, preferences, communication style)
- Internal POVs, playbooks, templates related to the topic
- Previous successful pitches or meeting outcomes on similar topics
- Usage: Leverage institutional knowledge, ensure consistency with past recommendations
- 路径:(预设文件夹结构)
~/personal-notes/ - 搜索内容:
- 与该主题相关的文档
- 关于对接人的记录(互动情况、偏好、沟通风格)
- 与主题相关的内部观点、操作手册、模板
- 过往类似主题的成功提案或会议成果
- 用途:利用机构知识,确保与过往建议保持一致
Workflow
工作流程
Phase 1: Research Gathering
阶段1:研究收集
- Perform web search on to identify industry context, trends, competitive landscape
[Topic] + [Company] - Search LinkedIn for:
- Person's profile (background, roles, recent activity)
- Company page and recent news
- Search Account Box folder for:
- Account-level summaries and relationship history
- Previous communications, proposals, known challenges
- Search Personal Notes folder for:
- Topic-specific playbooks or POVs
- Notes on the Person
- Related past successes
- 针对进行网页搜索,识别行业背景、趋势、竞争格局
[主题] + [公司] - 在LinkedIn上搜索:
- 对接人的个人主页(背景、职位、近期活动)
- 公司主页和近期新闻
- 搜索客户账户文件夹:
- 客户账户级摘要和合作历史
- 过往沟通内容、提案、已知挑战
- 搜索个人笔记文件夹:
- 主题专属操作手册或观点
- 关于对接人的记录
- 相关的过往成功案例
Phase 2: Analysis & Synthesis
阶段2:分析与整合
- Synthesize findings across all sources into one coherent narrative
- Identify the client's likely context, challenges, or opportunities
- Craft a tailored recommendation (the PoV)
- Organize supporting evidence by category for Speaker Notes
- 将所有来源的发现整合成一个连贯的叙事
- 识别客户可能的背景、挑战或机遇
- 撰写定制化建议(即观点内容)
- 按类别组织演讲备注的支撑证据
Phase 3: Output Generation
阶段3:输出生成
- Write Point of View section
- Write Speaker Notes with subsections and evidence
- Compile Sources section
- Format as clean, readable markdown
- 撰写观点章节
- 撰写包含子章节和证据的演讲备注
- 整理来源章节
- 格式化为清晰易读的Markdown文档
Quality Criteria
质量标准
Point of View
观点
- Tailored to the specific client and company (not generic)
- Addresses the stated topic
- Offers a clear recommendation or perspective
- Evidence-grounded in research findings
- Actionable (manager could deliver this in a meeting)
- Length: 200-400 words
- 针对特定客户和公司定制(非通用内容)
- 紧扣既定主题
- 提供清晰的建议或视角
- 基于研究结果,有证据支撑
- 可落地(经理可直接在会议中使用)
- 长度:200-400字
Speaker Notes
演讲备注
- Organized into logical subsections
- Detailed but scannable (paragraphs, not dense walls of text)
- Includes examples, data, or case studies where relevant
- Addresses anticipated questions or objections
- Grounded in research with clear sourcing
- Appropriate read time (5-7 minutes at ~250 WPM = ~1500 words max)
- 按逻辑组织成子章节
- 内容详细但易于扫描(段落形式,非密集大段文字)
- 相关处包含示例、数据或案例研究
- 涵盖预期问题或异议
- 基于研究内容,来源清晰
- 阅读时长合适(按每分钟250词计算,5-7分钟阅读对应最多约1500词)
Overall
整体要求
- Tone matches business context (professional, confident, not salesy)
- Free of generic platitudes or industry jargon without explanation
- Sources clearly attributed or noted
- Markdown is clean and easy to read
- 语气符合商务场景(专业、自信,非推销式)
- 无通用套话或未解释的行业术语
- 来源清晰标注
- Markdown格式清晰易读
Error Handling & Constraints
错误处理与约束
If Research is Limited
若研究内容有限
- Proceed with available sources (e.g., if Account Box is empty, lean on web + personal notes)
- Flag in output which sources were unavailable: "Note: Limited account history available; recommendations based on public research and industry best practices"
- 基于可用来源继续推进(例如:若客户账户文件夹为空,则依赖网页+个人笔记)
- 在输出中标记不可用的来源:"注意:可用客户历史记录有限;建议基于公开研究和行业最佳实践"
If Topic is Too Broad
若主题过于宽泛
- Ask user to clarify: "Is this meeting about [Specific Challenge]? That would help tailor the recommendation."
- 请求用户澄清:"本次会议是否围绕[具体挑战]展开?这将有助于定制建议内容。"
If Company/Person is Not Found
若未找到公司/对接人
- Proceed with general industry research
- Note: "Limited company-specific research available; PoV based on industry context"
- 基于通用行业研究继续推进
- 标注:"可用公司专属研究内容有限;观点基于行业背景"
Data Privacy
数据隐私
- Do not include confidential internal information (financials, strategic plans) in the markdown output
- Focus on synthesized, shareable insights
- Markdown输出中不得包含内部机密信息(财务数据、战略计划)
- 聚焦于整合后的、可分享的洞察
Output Example Structure
输出示例结构
markdown
undefinedmarkdown
undefinedClient Meeting Prep: [Person Name] @ [Company] | [Topic]
客户会议准备:[对接人姓名] @ [公司] | [主题]
Meeting Context: [Date if provided]
Prepared: [Current date]
Prepared: [Current date]
会议背景:[若提供则填写日期]
生成时间:[当前日期]
生成时间:[当前日期]
Point of View
观点
[2-3 paragraph narrative addressing the topic and client's specific situation, with a clear recommendation]
[2-3段叙事文本,围绕主题和客户具体情况展开,提供清晰建议]
Speaker Notes
演讲备注
Market Context
市场背景
[Context about the industry/trend]
[行业/趋势相关背景]
Client Situation
客户情况
[What you know about the company and person from research]
[从研究中了解到的公司和对接人信息]
Our Perspective / Value
我方视角/价值
[How to position your recommendation]
[如何定位我方建议]
Key Points to Cover
核心讨论点
[Talking points with evidence]
[有证据支撑的谈话要点]
Anticipated Objections
预期异议
[FAQ or likely concerns and responses]
[常见问题或可能的顾虑及回应]
Next Steps
后续步骤
[How to move forward from the meeting]
[会议后的推进方向]
Sources Consulted
参考来源
- [Source 1] (description)
- [Source 2] (description)
- ...
---- [来源1](描述)
- [来源2](描述)
- ...
---Configuration & Parameters
配置与参数
Environment
环境要求
- Requires: Web search capability, file system access (Account Box + Personal Notes)
- Optional: LinkedIn API or web scraping for social profile research
- 必需:网页搜索能力、文件系统访问权限(客户账户文件夹+个人笔记文件夹)
- 可选:LinkedIn API或网页抓取工具,用于社交主页研究
Execution
执行细节
- Estimated Runtime: 3-5 minutes for typical research + synthesis
- Output Format: Markdown (plain text, easily copied/pasted into docs, email, etc.)
- Output Length: Flexible; default 5-7 minute read time (~1500 words total)
- 预计运行时间:典型研究+整合需3-5分钟
- 输出格式:Markdown(纯文本,可轻松复制粘贴到文档、邮件等)
- 输出长度:灵活可调;默认阅读时长5-7分钟(总字数约1500词)
Known Limitations
已知限制
-
File System Access: Assumes specific folder structure (,
~/account-box/). If folder structure differs, user should specify paths.~/personal-notes/ -
LinkedIn Access: Web-based LinkedIn search is limited; deep API access may be restricted. Falls back to what's publicly available.
-
Proprietary Data: Cannot access password-protected databases or proprietary knowledge systems beyond the file system.
-
Real-Time Data: Web search reflects current publicly available information; may miss very recent announcements or internal company changes.
-
Confidentiality: User responsible for ensuring output doesn't contain sensitive internal information before sharing externally.
-
文件系统访问:依赖预设文件夹结构(、
~/account-box/)。若文件夹结构不同,用户需指定路径。~/personal-notes/ -
LinkedIn访问:基于网页的LinkedIn搜索存在限制;深度API访问可能受权限约束。 fallback方案是使用公开可得的信息。
-
专有数据:无法访问密码保护的数据库或超出文件系统范围的专有知识系统。
-
实时数据:网页搜索反映当前公开可得的信息;可能遗漏非常新的公告或公司内部变动。
-
保密性:用户需确保输出内容在分享前不包含敏感内部信息。
Success Metrics
未来优化方向
- Adoption: Manager uses the PoV directly in the meeting or as a starting point
- Relevance: Person/Company-specific details in the output (not generic)
- Actionability: Manager feels prepared and confident entering the meeting
- Time Savings: Reduces research and prep time vs. manual gathering
- 与CRM系统集成,获取结构化客户数据
- 实时LinkedIn集成,获取最新主页更新
- 基于历史会议笔记的AI驱动异议预测
- 主题专属观点模板库(行业操作手册)
- 基于公司垂直领域的自动化竞品分析
Future Enhancements
—
- Integration with CRM systems for structured account data
- Real-time LinkedIn integration for live profile updates
- AI-powered objection prediction based on historical meeting notes
- Template library for topic-specific POVs (industry playbooks)
- Automated competitor analysis based on company vertical
—