量身 AI← 案例库← Case library
案例 · AI 落地方案CASE · AI DEPLOYMENT PLAN · 财产险 · 总部滑铁卢,多伦多有大分支Property and casualty insurance · Waterloo, large Toronto presence

Definity 金融Definity Financial

收购 Travelers 之后,两家公司用同一把尺One ruler for two insurers, after the Travelers deal
基于公开资料的分析,与该企业没有合作关系。Based on public information; no affiliation with the company.

英文旁白、无字幕的 3–4 分钟管理层简报。3–4 minute executive brief, English narration, no captions. 在 YouTube 打开Open on YouTube ↗

一句话判断The verdict 成果是在一套业务上量出来的;合并后的公司需要统一口径来对比。Wins were measured on one book; the merged company needs one measure to compare them.
起步顺序Order of work 先统一度量口径,再统一理赔与核保,最后深化 Sonnet 的定价与反欺诈。Fix the measures first, then harmonise claims and underwriting, then deepen pricing and fraud on Sonnet.

公开事实Public facts 已核对Checked

  • 250+
    AI 与数据从业者AI and data practitioners2026 年 Q3 财报电话会Q3 2026 earnings call
  • 140+
    在运行的模型与解决方案Models and solutions in production已使用近 10 年Used for about ten years
  • 15%+
    客服与理赔生产力提升Productivity gain in service and claims同一电话会披露Reported on the same call
  • C$3.3B
    收购 Travelers Canada,2026 年 1 月 2 日完成Travelers Canada acquisition, closed 2 Jan 2026进一步扩大多伦多布局Adds a Toronto presence

两套业务,一把尺One ruler for both books 判断Judgment

  • 首次报案First notice of loss:自有业务工作量降约 50%Workload cut about 50% on its own book(Travelers 理赔也一样吗?Same result on Travelers claims?)
  • 客服与理赔Service and claims:生产力提升超 15%Productivity up over 15%(对照哪个基线?Against which baseline?)
  • 核保接收Underwriting intake:商业险已试点自动化Automation tried in commercial lines(报价更快是否带来保费?Does faster quoting lift premium?)

专家组裁决Expert panel 依据公开观点的推演Inference from public views

Thomas Davenport

整家企业怎么变Whole-enterprise change

并购是流程改变;每个统一的流程要有负责人。A merger is a process change; name the owner for each harmonised process.

Foteini Agrafioti

技术与业务战略对齐Strategy alignment

平台、Sonnet 与两家理赔团队的权责写成一页纸。Write one page on who owns what: platform, Sonnet and both claims teams.

Erik Brynjolfsson

价值与度量Value & measurement

两套业务用同一基线;整合培训算作成本。Use one baseline for both books; count integration training as cost.

Ethan Mollick

人怎么用How people use it

用 Travelers 的文件测试受理智能体;边界可能不同。Test the intake agent on Travelers files; edges may differ.

Ziad Obermeyer

算法在优化什么What it optimizes

Sonnet 的定价与反欺诈:写清真实目标与代理指标。For pricing and fraud on Sonnet, name the real goal and the proxy.

Andrew Ng

节奏与方法Pace & method

选一个理赔任务,三到六个月内在两套业务上跑出数字。Choose one claims task to run on both books in three to six months.

保留的分歧Disagreement kept · 抓住整合窗口 vs 先审定价Move in the window vs. review pricing

Andrew Ng:利用整合窗口快速上线。Use the integration window to ship fast.

Ziad Obermeyer:定价与反欺诈先过目标检查。Pricing and fraud pass a target check first.

→ 理赔任务走快速通道;定价与反欺诈完整审查。Claims tasks go fast; pricing and fraud get full review.

六位专家是谁、怎样思考、有什么铁律 →Who the six experts are, how they think, their rules →

三个目标Three priorities 建议Recommendation

1

统一一把度量尺Agree one measuring ruler

两套业务共用的定义与基线。Common definitions and baselines for both books.

2

先统一理赔Harmonise claims first

在 Travelers 理赔上运行首次报案工具并设对照组。Run the first-notice tool on Travelers claims with a control group.

3

在审查下深化 SonnetDeepen Sonnet with review

直销渠道的定价与反欺诈,先过目标检查。Pricing and fraud on the direct channel, after a target check.

场景组合Scenarios 建议Recommendation

场景Scenario做法Approach状态Status
Travelers 理赔统一Travelers claims harmonisation
提效Efficiency体验Experience
把两家公司的理赔处理统一到同一流程和口径。Bring both companies' claims handling onto one process and one measure.建议 · 整合正在进行Proposed · integration is under way
Sonnet 定价与反欺诈Sonnet pricing and fraud
增收Revenue降风险Risk
深化直销渠道的定价细分与文件篡改识别。Deepen pricing segmentation and document-tampering checks on the direct channel.建议 · 完整审查Proposed · full review
商业核保智能体Commercial underwriting agent
增收Revenue提效Efficiency
规模化申请接收,让报价更快;由人决定。Scale submission intake so quotes go out faster; a person decides.建议 · 基于商业险试点Proposed · builds on a commercial pilot
统一客服平台One contact-centre stack
提效Efficiency体验Experience
两家公司的客服共用已部署的 Google Cloud 工具。Run both companies' service on the Google Cloud tools already in place.建议 · 基于已上线工具Proposed · builds on live tools

12 个月路线12-month roadmap 建议Recommendation

奠基Foundation

第 1–3 月Months 1–3
  • 统一定义One set of definitions
  • 两套业务基线Baselines for both books
  • 模型清单与分级Model inventory, tiers
◆ 度量口径确认Ruler agreed

试点Pilot

第 4–6 月Months 4–6
  • 在 Travelers 运行报案工具First-notice tool on Travelers
  • 对照组Control group
  • 边界案例记录Edge-case log
◆ 指标达成Targets met

上线Go live

第 7–9 月Months 7–9
  • 理赔统一流程Claims on one process
  • 关键步骤人工确认Person confirms key steps
  • 上线即监控Monitoring from day one
◆ 风险合规签字Risk sign-off

扩展Extend

第 10–12 月Months 10–12
  • Sonnet 定价评审Sonnet pricing review
  • 商业核保智能体Commercial underwriting agent
  • 年度重排优先级Annual re-ranking
◆ 进入下一年Next year begins

治理与合规Governance 建议Recommendation

  • 模型清单对齐 OSFI E-23(2027-05-01 生效);适用性请法务确认Model inventory aligned to OSFI E-23 (effective 2027-05-01); confirm applicability with counsel
  • 定价与反欺诈模型:完整审查Pricing and fraud models: full review
  • 核保与理赔决定由人确认A person confirms underwriting and claims decisions

三层指标Three-layer metrics 建议Recommendation

采用Adoption每周使用工具的理赔人员Claims handlers using the tool weekly · 走统一流程的理赔占比Share of claims on the shared process
过程Process两套业务的处理时长Handling time on both books · 人工确认通过率Human confirmation pass rate · 报价周转时间Quote turnaround
结果Outcome单笔理赔成本对照基线Cost per claim against baseline · 更快报价带来的保费Premium won from faster quotes

目标值由企业自定。Targets are the company's to set.

风险Risks 判断Judgment

两套口径,两种故事Two measures, two stories

如果两套业务各算各的,成果无法比较。If each book counts wins its own way, results cannot be compared.

整合负荷过重Integration overload

正在合并的团队可能没有余力接新工具;排好顺序。Teams already merging may have no room for new tools; sequence carefully.

Sonnet 定价中的代理指标Pricing proxies on Sonnet

只优化转化率会掩盖风险;按客户群检查。Optimising for quick conversion can hide risk; check by customer group.

前 90 天First 90 days 建议Recommendation

第 1–2 周Weeks 1–2点名度量口径负责人Name the measurement owner
第 3–4 周Weeks 3–4盘点两家公司的模型Inventory models on both companies
第 5–10 周Weeks 5–10确认基线;选定一个理赔任务Agree baselines; pick one claims task
第 13 周Week 13首次复盘First review
今天只需要一个决定:One decision today: 为共用度量口径指定一位负责人name one owner for the shared measuring ruler

我们可能错在:We could be wrong if: Travelers 的理赔差异太大,首次报案的成果无法迁移。Travelers claims differ so much that the first-notice result does not carry across.

本页是基于公开资料的分析,与所述企业没有合作关系,不代表其官方立场或内部计划;企业数字多来自其自述或媒体转述;「专家组」是依据六位专家公开观点的推演,不是他们的原话,也不代表他们或其雇主的立场。资料整理日期:2026-11-14。This page is an analysis based on public information. It has no affiliation with the company and does not represent its official position or internal plans. Company figures come mostly from its own statements or media reports. The expert panel is an inference from six experts' public views, not their words, and does not represent them or their employers. Compiled 2026-11-14.

想给你的公司也做一份?Want one for your company?

同样的方法:公开事实、专属判断、12 个月路线和前 90 天。公司大小都可以聊,中文沟通。Same method: public facts, a company-specific judgment, a 12-month roadmap and the first 90 days. Any size of business; we can talk in Chinese or English.

Telegram @johnsunlei邮件 Email [email protected]