Software Engineer - Model Evaluation and Productisation (Must be UK based)

PolyAI is a leader in automating customer service through innovative voice technology. Our voice assistants empower businesses to deliver exceptional customer service at every interaction.

We are seeking a talented and hands-on Software Engineer with a strong data science background to join our team.

In this role, you will work on building software to enhance the visibility and configurability of large language models (LLMs). You will be responsible for rapidly developing tools and platforms to evaluate, iterate, and productionalize models, ensuring their reliability and accuracy.

We are looking for the right candidate, and therefore are flexible on the levelling for this position ranging from mid- level to senior!

Responsibilities:

• Must have at least 2 years of Python experience
• Must have at least 2+ years working experience
• Develop software that provides visibility into LLM models and offers configurability for tuning and evaluation.
• Build and maintain evaluation datasets and tools, enabling the measurement of model performance across key metrics.
• Take a hands-on approach to quickly prototype, test, and iterate on solutions, moving from proof-of-concept to production.
• Employ a data-driven methodology to drive model accuracy, leveraging evaluation results to inform decisions.
• Collaborate with cross-functional teams to integrate developed tools and ensure they meet production standards.
• Formulate hypotheses, design experiments, and collect data to validate model assumptions, consistently striving for improved reliability.
• Communicate findings and ranking metrics clearly to both technical and non-technical stakeholders.

Why Join Us:

Join a dynamic and innovative team at the forefront of LLM development. You will have the opportunity to work on challenging projects, rapidly build impactful solutions, and drive data-informed improvements that push the boundaries of what LLMs can achieve.

• Degree in Computer Science, Data Science, or related field, or equivalent experience.
• Strong proficiency in Python
• Experience developing evaluation suites, datasets, and data-driven tools for model reliability testing.
• Ability to rapidly prototype and iterate on solutions while maintaining a focus on production-level quality.
• Strong problem-solving skills and a creative mindset, with the ability to hypothesize and validate results through experimentation.
• Familiarity with cloud platforms such as AWS, GCP, or Azure is a plus.