40%
Professional experience
TuSimple — Freelance Data Analytics
This was highly practical analytics engineering work: automate the repetitive, improve the slow, and reduce the chance that downstream analytical models inherit bad inputs.
40%
30%
ETL + SQL
Overview
Automated ETL pipelines, embedded validation checks and faster reporting on large operational datasets.
Operational reporting pipelines often fail quietly through bad inputs, slow queries and repetitive manual steps. The challenge here was to make the system both faster and more trustworthy.
01
Automated recurring reporting workflows with Python-based ETL design.
02
Added data validation and anomaly detection checks so quality issues surfaced earlier in the pipeline.
03
Optimized SQL queries across large-scale datasets to reduce reporting latency.
04
Applied data wrangling processes that improved input reliability for downstream models and analyses.
01
Reduced manual reporting work by 40%.
02
Improved reporting efficiency by 30%.
03
Created cleaner, safer upstream data conditions for analytics consumption.
Detailed notes
What the work demonstrates.
Automation focus
The project was centered on eliminating repetitive reporting effort by moving logic into reproducible ETL workflows.
Data trust
Embedded validation and anomaly checks improved trust in inputs before they reached downstream analysis layers.
Performance improvement
SQL optimization on large datasets materially improved turnaround time, helping reporting become more responsive to business needs.
Engineering maturity
The work highlights readiness for analytics engineering and data engineering contexts where correctness and efficiency must coexist.