About Me
I'm a passionate software engineer and UCL Computer Science Master's student (graduating in 2025) with experience in backend development, AI, and building scalable tools. I've worked on automating trading processes at Macquarie, enhancing platform observability at Thought Machine, and developing AI-driven tools for marketers as an early hire at Kaya. My work focuses on creating impactful, efficient solutions, and I'm driven by a love for solving real-world problems and advancing technology.
Experience
Software Engineer
Kaya (YC 21) | Oct 2024 - Present
- Improved search speed by implementing Kaya Search, a vector-based search system optimized for marketers' needs.
- Developed an AI chatbot agent to help marketers generate ad content and ad copies efficiently.
- One of the first engineering hires.
Software Engineer
Macquarie Group | June 2024 - Aug 2024
- Automated daily risk team emails and decommissioned an external data platform using Python Airflow, saving over £2000 per month.
- Streamlined trading data ingestion by implementing concurrent API calls, reducing process time from 1.5 hours to 0.5 hours and enhancing efficiency by 200%.
- Pioneered the automation of database schema updates.
Backend Engineer
Thought Machine | July 2023 - Sept 2023
- Enhanced Deployment Hub API observability by integrating Prometheus metrics in Golang and configuring kube-state-metrics.
- Created interactive Grafana dashboards for Kubernetes metrics visualization and improved user experience with a React.js frontend.
- Implemented database restoration metrics to optimize system monitoring and troubleshooting.
Machine Learning Contributor
Unify AI / Ivy | Jan 2024 - Dec 2024
- Contributed to the Ivy unified machine learning library by fixing critical test cases and implementing new features to improve compatibility across frameworks like TensorFlow and PyTorch.
- Enhanced library robustness by identifying and resolving edge cases, ensuring seamless performance in real-world applications.
- Collaborated with the open-source community to review and refine pull requests, promoting high coding standards and maintainability.
- Improved the developer experience by documenting features and providing clear examples for new functionalities.
Projects
Energy-Efficient Stable Diffusion Framework
Investigated the impact of energy consumption on image quality in Stable Diffusion models using Genetic Algorithms (NSGA2 and NSGA3) and CodeCarbon. Analyzed how CPU, GPU, RAM, and runtime affect image quality, optimizing heuristics to balance energy efficiency and output quality effectively.
Bias Mitigation in AI-Generated STEM Images
Optimized Stable Diffusion using Genetic Algorithms to improve representation, achieving a 20% increase in female representation and balanced skin tone distribution.
Personal Linter
Developed a custom linter using Abstract Syntax Trees (AST) to enhance code quality, shared for peer use.
Cascading Vulnerabilities in Python Package Ecosystem
Conducted a network analysis of the Python Package Index (PyPI) to study its scale-free properties and vulnerability to cascading failures. Explored structural patterns, rich club formation, and degree correlations to identify critical packages and propose strategies to enhance the ecosystem’s resilience.
Related Work
Open-Source Contributions to Ivy
Improved the Ivy machine learning framework by fixing test cases, implementing new features, and enhancing compatibility across various systems. Contributed to the development of Ivy’s unified machine learning library to simplify the use of multiple backends like TensorFlow and PyTorch.
Learn MoreComplex Networks Analysis of PyPI Ecosystem
Conducted a comprehensive network analysis of the Python Package Index (PyPI) to examine its scale-free properties and cascading failure vulnerabilities. Proposed strategies for improving ecosystem resilience by analyzing rich club formation and dependency correlations.
Learn MoreBias Mitigation in Stable Diffusion
Investigated bias in AI-generated STEM images by optimizing Stable Diffusion models with Genetic Algorithms. Achieved a 20% increase in female representation and a balanced skin tone distribution, awarded the UCL Computer Science EDI Award.
Learn More