📌 Note: This repository serves as a summary and entry point. The fully executed code, interactive data frames, and complete analysis are hosted on Kaggle.
Working as a data consultant for an international HR firm, this project analyzes a comprehensive dataset of over 57,000 global tech salary records (2020-2024). The goal is to uncover macro-level insights into global compensation drivers, the evolving reality of remote work, and hiring strategies across more than 200 unique tech roles.
- Engine/Environment: SQL (PostgreSQL syntax via DuckDB) integrated into a Kaggle Python (Pandas) environment.
- Advanced Implementations: Multi-level Common Table Expressions (CTEs), Window Functions (
DENSE_RANK,LAG), Conditional Logic/Aggregations (CASE WHEN), and Data Validation/Skepticism analysis.
Here is a glimpse of the strategic findings uncovered during the query execution:
- The "Autonomy" Premium: Moving from Entry-level to Mid-level yields the highest relative salary jump in a tech professional's career (+27.38%), signaling high market valuation for independent execution.
- The Corporate Premium Gap: Large enterprises pay an average of 30% more than startups for equivalent experience. This financial gap severely peaks at the Mid-level tier, where startups pay nearly 49% less.
- The Remote Work Paradox: Non-executive tiers experience a "remote discount," accepting slightly lower salaries for full-remote roles compared to on-site. Executive roles completely invert this trend, commanding top-dollar packages ($216K+ average) while working 100% remotely.
- Infrastructure Valuation: Specialized roles in AI, Machine Learning, and Site Reliability Engineering (SRE) command the highest market compensation, highlighting the premium placed on platform stability when scaling AI systems.
README.md-> Executive project brief and summary of results.notebooks/tech-salary-analysis.ipynb-> Jupyter notebook from Kaggle for visualization
- LinkedIn: linkedin.com/in/seu-perfil
- Full Portfolio & Notebook: Kaggle Profile
Inspired by an initial dataset from DataCamp and independently scaled into an advanced data consulting case study.