Kendi Ng Ka Wai

Marketing Analytics & Digital Strategist | Python & AI Integration | China Media | Hong Kong

I build data systems that turn raw marketing signals into business decisions — and I have spent 15 years understanding the platforms those signals come from. My background spans China media platforms (Baidu, Tencent, Douyin, RED) and international channels (Google, Meta, LinkedIn), and I have built my own analytics stack on top: Python ETL pipelines, PostgreSQL, self-built Streamlit dashboards, and production Claude AI API integration. I am looking for in-house roles in Hong Kong where data, marketing, and technology intersect.

View Projects LinkedIn GitHub Download CV

Skills & Technologies

Data Engineering

Python
ETL Pipelines
PostgreSQL
SQL
Pandas
NumPy
VBA Excel
GitHub

Analytics & Visualisation

Streamlit
Tableau
Power BI
Looker Studio
matplotlib
seaborn
GA4
Google Search Console

Marketing Technology

Google Ads
LinkedIn Ads
LinkedIn Marketing API
GA4
GTM
Claude AI API
Azure Graph API
Google Apps Script

China & APAC Platforms

Baidu Search & Display
WeChat Moments Ads
Tencent Video
Douyin
RED / Xiaohongshu

CRM & Business Analysis

Odoo CRM
Zoho CRM
Jira
UAT
Requirements Gathering
Data Mapping

Web Frameworks

Streamlit
FastAPI
HTML / CSS
JavaScript

Automation & DevOps

Docker
AWS EC2
APScheduler
Windows Task Scheduler
Microsoft Graph API
Outlook 365 Automation
BeautifulSoup

Featured Projects

All projects below are part of one integrated analytics platform I built from scratch. Each card links to the relevant code within the main repository.

Digital Marketing Analytics Platform

Problem: Marketing data was scattered across SEO, LinkedIn Ads, LinkedIn Lead Forms, and GA4 with no unified analytics view or AI-powered insights.

Built a unified digital marketing analytics hub with 4 core modules: SEO keyword tracking (Brand vs. Paid segments with MoM/YoY comparisons), LinkedIn Lead Gen Form sync via API, LinkedIn Agency report ingestion from Excel uploads, and Digital Marketing campaign drill-down from Odoo CRM. Integrated Claude AI API for automated executive summary generation and insight buttons throughout.

Python
Streamlit
PostgreSQL
Search Console API
LinkedIn Marketing API
GA4 Data API
Claude AI API
Pandas

GitHub

CRM Reporting & Analytics Dashboard

Problem: A financial services company had no internal BI tool — sales pipeline, pension AUM, and marketing data lived in disconnected CRM tables with no interactive reporting.

Built a multi-page Streamlit BI dashboard serving 8+ interactive pages: EBCAS sales pipeline analytics by product team with fiscal year filtering, PIES pension AUM analysis with multi-scheme classification and MPF provider breakdown, and marketing campaign performance reporting. Includes automated email report generation via Microsoft Graph API.

Python
Streamlit
PostgreSQL
Pandas
Docker
Microsoft Graph API

GitHub

Data Warehouse & ETL Pipeline

Problem: Marketing and sales data from Google Search Console, GA4, and LinkedIn APIs had no centralised storage, making cross-source analysis impossible.

Designed and built a local PostgreSQL data warehouse with normalised tables for SEO reports, keyword data, LinkedIn leads, agency data, and GA4 leads. Created Tableau-ready SQL views (unified monthly summary, agency targeting breakdown, audience demographics). Built 4 automated ETL scripts pulling from Search Console API, GA4 Data API, LinkedIn Lead Gen API, and LinkedIn Ads Reporting API into PostgreSQL on a scheduled basis.

PostgreSQL
Python
Search Console API
GA4 API
LinkedIn Marketing API
Google Sheets API
Tableau
psycopg2

GitHub

Automated Email Report System

Problem: 10+ recurring business reports (sales insights, pension reports, campaign summaries) were being compiled and sent manually, consuming hours of analyst time.

Built a comprehensive automated email reporting engine with HTML-formatted reports, styled data tables with conditional formatting, and Excel attachments — all sent via Microsoft Graph API. Powered by APScheduler with configurable cron schedules via JSON, replacing Windows Task Scheduler. Report types include Combined Sales Insight, Individual Sales, EBCAS, Pension, Marketing Campaign, and provider-specific reports.

Python
APScheduler
Microsoft Graph API
MSAL
Pandas
HTML/CSS Email

GitHub

AI-Powered SEO Reporting Pipeline

Problem: Monthly SEO performance reporting required hours of manual analysis and formatting before it could be sent to senior management.

Built a Google Apps Script pipeline that connects to Search Console API, tracks 3 keyword segments (Brand, Paid Maintain, Paid New), generates MoM/YoY performance comparisons, and sends HTML email reports with monthly trend tables. Integrated Claude AI API for automated executive summary narratives. Reports distributed to 9 stakeholders via Microsoft Azure Graph API — zero manual intervention.

Google Apps Script
Claude AI API
Search Console API
Azure Graph API
HTML
JavaScript

GitHub

More Projects

Lead Management & CRM Automation

Built lead assignment (10 preset workflows), lead release, and contact analysis tools that automate CRM operations and generate Odoo-importable files. Includes 7-stage filter pipeline for data quality monitoring.

Python
Streamlit
PostgreSQL
Pandas

GitHub

Sales Data Processing & Reporting Engine

Comprehensive sales data pipeline processing CRM exports across 6 report modules: opportunity pipeline, activity tracking, customer lead conversion, target vs. actual performance, and executive Combined Sales Insight dashboard.

Python
Pandas
Odoo CRM
openpyxl

GitHub

Data Comparison & Quality Tools

FastAPI web portal for internal data comparison tools. Lead vs. Sales Rep comparison engine with 4 exclusion rules, live schedule viewer for automated tasks, and card-based navigation. Dockerised deployment.

FastAPI
Python
Docker
Bootstrap 5

GitHub

Company Name Matching Algorithm

Fuzzy company name matching tool using sequence matching and Unicode normalisation for CRM data deduplication. Handles bilingual names (English/Chinese) and cross-platform path handling.

Python
difflib
Pandas
Unicode

GitHub

Odoo CRM Data Export Suite

6 automated data export scripts pulling from Odoo CRM's PostgreSQL database — activities (59 columns with auto header mapping), contacts, leads, opportunities, sales rep logs, and AUM detail.

Python
psycopg2
PostgreSQL
Odoo ERP

GitHub

Monthly GUMbees TV & Won Case Reports

Automated monthly internal reports with data visualisation charts embedded directly in emails. Won Case quarterly report generates 3 heatmaps + 1 stacked bar chart as inline CID images in HTML emails.

Python
matplotlib
seaborn
Graph API

GitHub

Weekly Data Quality Monitoring

Automated weekly monitoring that detects dissolved companies and parent-subsidiary mismatches in CRM data. Sends alert emails with 4 filter categories and automated correction of parent/subsidiary files.

Python
Pandas
Graph API
PostgreSQL

GitHub

Docker Containerisation & Infrastructure

Containerised the entire reporting platform with Docker for AWS EC2 deployment. Multi-service Docker Compose with Streamlit + FastAPI, health checks, volume mounts, and environment-based configuration.

Docker
Docker Compose
AWS EC2
Linux

GitHub

Automated Job Application Tracker

Python automation that reads Outlook 365 emails from JobsDB via Microsoft Graph API, parses application status updates, and updates an Excel tracking file on OneDrive automatically. Runs hourly.

Python
Graph API
Azure AD
BeautifulSoup

GitHub

CRM Migration Planning (Freelance)

Led full Zoho CRM migration planning for a social enterprise as sole consultant — stakeholder interviews, module evaluation, and complete data mapping documentation.

Zoho CRM
Data Mapping
Requirements
Consulting

Career Journey

2007

AXA

Finance

2010

Publicitas

APAC Publications

2015

iClick

China Media — Senior Director

2024

Gain Miles

MPF/Insurance — CRM & Analytics