Self-Improving AI Tooling Architect

Team Wheel Limited; Greater London; On-site; GBP 80,000 - 110,000; Full time

Benefits offered by this job

Shadow equity

Flexible working

Job summary

Team Wheel Limited is seeking an AI Tooling Specialist to build self-improving AI systems in London. You will design and implement recursive learning loops, automated evaluation, and memory mechanisms, taking ownership from architecture to production.

Ideal candidates have strong Python experience with LLMs, experience designing iterative improvement loops, and familiarity with evaluation design and prompt/chain optimisation.

Qualifications

• Strong Python and hands-on experience building with LLMs at a production level.
• Demonstrable experience designing recursive or iterative self-improvement loops in AI systems.
• Deep familiarity with evaluation design — automated scoring, preference modelling, critic models.
• Experience with prompt/chain optimisation frameworks (DSPy, TextGrad, or equivalent).
• Solid understanding of agent architectures and multi-step reasoning pipelines.
• Ability to articulate why a self-learning loop failed, not just that it did.

Responsibilities

• Design recursive learning loops where AI outputs are scored and fed back into generations.
• Build self-critique and self-refinement pipelines for improved responses before delivery.
• Implement Constitutional AI-style approaches to drive automated improvement without human-in-the-loop on every iteration.
• Design memory and state mechanisms to apply learnings across sessions and tasks.
• Develop automated evaluation frameworks (evals) and feedback signal pipelines.

Skills

Python programming

LLMs in production

Recursive self-improvement

Evaluation design

Prompt/chain optimisation

Agent architectures

Troubleshooting self-learning loops

Tools

DSPy

TextGrad

Job description

Team Wheel Limited is seeking an AI Tooling Specialist to build self-improving AI systems in London. You will design and implement recursive learning loops, automated evaluation, and memory mechanisms, taking ownership from architecture to production.

Ideal candidates have strong Python experience with LLMs, experience designing iterative improvement loops, and familiarity with evaluation design and prompt/chain optimisation.