Kai Xiao
Senior Consultant
Kai Xiao
About Me
I work at the intersection of business analysis, Microsoft Dynamics 365 delivery, and practical AI product building. My strength is translating messy operational needs into clear requirements, usable workflows, and systems that teams can actually adopt.
This site is both a personal portfolio and a small AI project hub. It shows how I use AI not as a buzzword, but as a delivery layer: prompt design, document generation, CRM process support, cost controls, and deployment thinking.
- LocationCanada
- FocusCRM, D365, AI workflows
- MarketEnglish-speaking business users
- PortfolioAI projects + consulting cases
- LinkedInkaixiaocan
What I Do
Resume
Experience
Current
Consulting & AI ProjectsSenior Consultant / AI Project Builder
Designing practical AI workflows, local and cloud deployment models, and business-facing project demos that connect AI output with real user tasks.
Enterprise
CRM / ERP TransformationFunctional Consultant
Supported Dynamics 365 and cloud migration projects by translating business process gaps into solution requirements, test scenarios, and implementation priorities.
Public Sector
D365 Cloud MigrationBusiness Analyst / Functional Consultant
Helped connect legacy CRM requirements, security considerations, plugin/workflow modernization, and UAT coordination for a large user base.
Education & Learning
Ongoing
Professional DevelopmentAI, CRM, Power Platform
Continuous learning around Dynamics 365, Copilot-style business workflows, Cloudflare deployment, and AI-assisted productivity tools.
Delivery Skills
Business Analysis
Dynamics 365 / CRM
UAT & Stakeholder Alignment
Documentation
Technical Skills
AI Prompt & Workflow Design
Python / API Integration
Cloudflare / Docker / Nginx
Knowledge
- Dynamics 365
- CRM
- Power Platform
- Business Central
- AI Workflows
- Prompt Design
- Cloudflare
- Docker
- UAT
- BI Thinking
- Documentation
Selected Projects
Career Prep Platform
AI ProductJD + resume input becomes a targeted prep pack with resume notes, company brief, and interview Q&A.
Public-Sector CRM Migration
Dynamics 365Migrating legacy CRM logic into a cleaner Microsoft cloud operating model.
Enterprise Digital Transformation
CRM + ERPModernizing disconnected sales, service, finance, warehouse, and BI workflows.
FNOS Local Web Server
Cloud / NASDocker, Nginx, Python service, Cloudflare Tunnel, and static site hosting on local NAS.
Industry Insights
Notes from a practical delivery angle: what changes for business users, what becomes easier to automate, and what still needs careful process design.
AI in CRM works best after the process is clear.
Copilot-style features can summarize, route, and recommend. They still need clean ownership, reliable data, and agreed escalation rules.
Contact centers are moving closer to knowledge management.
Service work is no longer only about answering cases. It is about connecting profile, history, knowledge, channels, and follow-up actions.
Low-code is strongest when governance is visible.
Power Apps and automation are useful because business teams can move faster. The tradeoff is needing naming, ownership, and environment discipline early.
A demo should show the workflow, not only the model.
Useful AI demos explain inputs, limits, outputs, cost controls, and where human review fits. That makes them easier for non-technical users to trust.
AI Work
Career Prep Platform
The first public demo is an AI-assisted career preparation workflow. It accepts a company, target role, region, job description, and resume, then produces a practical starter pack for English-market interview preparation.
The product thinking behind it matters: daily role limits, downloadable Markdown/TXT output, a 30-day retention model, and a clean separation between user-facing results and internal prompts.
Open Project Overview Try WorkflowInputs
JD + ResumeOutputs
4-PackGuardrails
3 / DayContact
How Can I Help?
I am interested in CRM modernization, Dynamics 365 delivery, AI-assisted business workflows, and practical prototypes that help teams understand a process faster.
For the quickest context, send the business problem, current system landscape, user group, and the decision you are trying to make.