Data & AI Engineer · Full-Stack Developer

Ding Hock Wei
makes ad spend accountable.

20+ years building PHP and Python systems. For the past ten years I've run the data platform at a US mass tort law firm: five ad and CRM sources synced every hour, LLMs that score every lead, ML models that rank every marketing partner, and AI bots that make the first call.

pipeline · hourly sync● DEMO
records today: 0syncing…
0
years shipping production systems
0
ad & CRM sources in one pipeline
0
syncs a day, one every hour
0
records processed per day
0
Metabase dashboards built
0
LLM families fine-tuned: Llama, Qwen, GLM
Industries I've built for

The pipeline I own

Every ad click, lead, intake call and signed case lands in one place, gets cleaned and matched, then feeds the dashboards and scoring models that marketers, partners and intake use every day.

Wilenslaw · 2016 – now

What the data shows

Examples of the reporting I build: ingestion volume, the click-to-case funnel, case mix by tort, the manual hours automation replaced, and the partner scores behind scaling decisions.

Illustrative data

Records ingested per hour

A typical day, all 5 sources · hover a bar

Case mix by tort

Share of signed cases

Click → signed case funnel

Monthly volume at each stage

Manual reporting hours per week

Before and after the automated pipeline

Marketing partner scorecard

ML score for each partner · bubble size = monthly spend · hover a partner
Score 75+ · scale up50–74 · optimiseUnder 50 · cut back

22 years, one timeline

Agency work, video platforms, MLM software and confidential contract work, then a decade in legal-marketing data and AI.

2004 – 2026

Skills

Strongest where data, integrations and AI meet, on top of two decades of running my own servers.

Self-assessed

Skill profile

0 – 100

Toolbox

Grouped by area

How my skills connect

Five skill areas and the tools that link them · hover a node

Experience