# Aayush Bajaj's CV
*Data Analyst*

- Phone: +61 481 910 408
- Email: [j@abaj.ai](mailto:j@abaj.ai)
- Location: Sydney
- Website: [abaj.ai](https://abaj.ai/)
- GitHub: [abaj8494](https://github.com/abaj8494)
- LinkedIn: [abaj](https://linkedin.com/in/abaj)


# Education

## University of New South Wales, Master in Statistics

- Feb 2026 – Dec 2028
- Sydney, Australia

## University of New South Wales, Bachelor in Computer Science (AI), Minor Mathematics

- Feb 2021 – Sept 2025
- Sydney, Australia

# Skills

- Statistical Analysis: Hypothesis Testing, Regression, Time Series, Experimental Design, A/B Testing, Bayesian Methods
- Data Engineering: SQL, PostgreSQL, ETL Pipelines, Data Warehousing, Azure SQL, pgvector
- Analytics & Visualisation: Python, Pandas, NumPy, Matplotlib, Seaborn, Statistical Modeling, Data Cleaning
- Cloud & Infrastructure: Azure, Docker, RESTful APIs, Data Pipeline Orchestration
- Development Tools: Git, Jupyter, Linux/Shell, Version Control, Automated Testing
# Projects

## [Biomedical Semantic Segmentation](https://abaj.ai/projects/dl/kits19)

- Oct 2024 – June 2025
- Implemented and compared 2D/3D state-of-the-art architectures; achieved class-leading Dice score of 78% (full marks).
- Reproduced nnU-Net end-to-end on Nvidia H200/A200; published technical report with ablation studies.
- Deployed to KiTS19 official leaderboard: ranked #57 globally with Dice score of 0.9129 through systematic hyperparameter optimisation.

## [Dead Tree Segmentation](https://abaj.ai/doc/pubs/dead-tree-seg.pdf)

- Kaggle
- Implemented U-Net from scratch (31M parameters); combined Dice + Focal loss for 97% class imbalance.
- Achieved 0.71 IoU after morphological post-processing; parallelized training on CUDA-enabled HPC.

## [10,000 Hours Analytics](https://github.com/abaj8494/10khrs-ai-ml-dl)

- Aug 2024 – present
- Comprehensive Jupyter notebooks for supervised/unsupervised learning with statistical rigor.
- Version-controlled experiments demonstrating data preprocessing, feature engineering, and model validation.

## [ABAJ.AI Infrastructure](https://abaj.ai)

- May 2023 – present
- Automated publishing pipeline built with Hugo; deployed behind Nginx with HTTPS.
- Comprehensive ML documentation following Zettelkasten methodology for knowledge graphs.

# Experience

## Freelance, Technical Consultant

- 2024 – present
- Sydney, NSW
- Deployed data infrastructure and analytics pipelines for client web applications.
- Configured database systems, ETL workflows, and monitoring dashboards for production environments.

# Certifications

## AWS Certified Machine Learning – Specialty


## Microsoft Certified: Azure AI Engineer Associate (AI-102)