Christopher Little-Savage, Engineering Lead

Prototyping AI agent systems to explore startup ideas including:

• Code Modernisation: Static analysis to deterministically translate legacy COBOL
• Coding Agent: Full-stack agent with file system manipulation and error correction.
• Code Review: Visualisations to help human reviewers debug agent-written code.
• Excel Automation: Custom execution environment to run AI-generated analyses.

As the founding engineer, I laid the foundations for the PowerX AI platform, establishing the patterns for distributed ML pipelines and high-frequency data processing. I maintained these critical infrastructure components while scaling my team to 15 people during this period of high growth.

As my responsibilies increased to include hiring, mentorship and technical roadmap ownership, I proudly remained deeply hands-on as the most active contributor to the codebase by all metrics. I worked closely with domain experts to translate complex power electronics requirements into robust software, bridging the gap from hardware to cloud.

Managing an international team of six external and internal developers, I was fully responsible for the delivery of technical solutions from research prototype through to released product. As well as this managerial work, I was also involved in all aspects of implementation, including solution design, development and code review.

Our flagship product was the GSMA Interoperability Test Platform, which aimed to harmonise APIs between some of the world's biggest network operators.

I started Tyro to tackle the technical skills gap in Uganda. I combined a unique curriculum of modern web development skills with a collaborative Agile workflow which allowed our apprentices to work on commercial projects with very limited practical experience.

I was directly involved in the recruitment of a team of nine local developers of mixed abilities, and was responsible for continuously monitoring their learning and development alongside making technical contributions on more advanced projects.

Using my expertise in rapid product development, I have been able to advise and assist in the delivery of proofs of concept and minimum viable products for startups around the world.

I took courses covering Machine Learning, Data-Processing Systems, Security, Functional Programming and Programming Language Usability. My final dissertation looked at building a hybrid graph database aiming to bridge the performance gap with relational databases.

For my undergraduate dissertation, I designed and implemented a novel type-inference system for JavaScript, and accompanying compiler to protect typed JavaScript from unchecked code. I went on to present this research at an international conference on functional programming.