AI/ML & Prompt Engineer LLM - RAG & Voice Agent
About the job
- Demonstrable hands-on experience with prompt engineering, large language models such as GPT, LLaMA, or Mistral, retrieval-augmented generation frameworks, and voice-agent or dialogue-system design, with shipped projects or open-source contributions.
- Advanced proficiency in Python for machine-learning and artificial-intelligence development.
- Experience building application programming interfaces in NestJS and familiarity with Next.js and React for frontend integration.
- Strong understanding of Azure services, including App Services, Functions, and Cognitive Services; containerisation with Docker; and relational databases such as MySQL.
- Experience designing microservices, distributed architectures, and RESTful or GraphQL APIs.
- Excellent written and verbal communication skills and the ability to articulate complex AI and voice-agent concepts to technical and non-technical stakeholders.
- Comfort working remotely and collaborating asynchronously with a UK-based team during UK business hours, 9:00 AM to 6:00 PM GMT/BST.
- Architect and deploy LLM- and RAG-based systems that ingest data sources, generate accurate responses, and power chatbot and voice-agent interactions.
- Design, iterate, and optimise prompts and voice-agent dialogue flows to maximise response relevance, reduce latency, and ensure clinical safety across text and voice channels.
- Partner with frontend teams using Next.js and React and backend teams using NestJS and Python to integrate AI and voice-agent components into the Azure-hosted microservices architecture.
- Instrument AI and voice-agent pipelines, analyse logs and user feedback, troubleshoot edge cases, and implement continuous-learning improvements.
- Stay current with advances in LLMs, RAG, conversational AI frameworks, GDPR, and DTAC to inform the technical roadmap.
- Define and execute rigorous test plans with QA to validate model accuracy, voice-agent performance, and compliance with healthcare standards.
- Mentor junior engineers and advocate for best practices in MLOps, prompt engineering, voice-agent design, and model governance.
- Contributions to open-source LLM or RAG tooling.
- Experience with MLOps pipelines, model monitoring, and automated evaluation frameworks.
- Background in speech-to-text, text-to-speech, or telephony integration.
About the company
EMMA by QuantumLoopAI is an AI receptionist designed for NHS primary care. It answers patient calls during GP surgery hours instantly, with no queue or hold music. EMMA handles both clinical and administrative inquiries, triages to the right outcome, and integrates with NHS tools like Accurx. It is used across hundreds of GP surgeries and PCNs, serving over 2 million NHS patients. This reduces call-handling costs (up to about 80%) and improves patient access and satisfaction, while easing pressure on reception staff. QuantumLoopAI emphasizes NHS clinical safety, DTAC certification, and alignment with NHS standards. The company’s goal is to make primary care more accessible, efficient, and sustainable by eliminating phone bottlenecks and letting reception teams focus on in-person patient care.
Company Size
51-200
Company Stage
N/A
Total Funding
N/A
Headquarters
London, United Kingdom
Founded
2023
Get referred to QuantumLoopAI
See people who can refer or advise you
Simplify's Take
What believers are saying
- QuantumLoopAI said on March 23, 2026 it raised over £6 million and serves more than 2 million patients.
- The company is listed on the Crown Commercial Service RM6200 framework, easing NHS procurement.
- QuantumLoopAI says practices save 15 workdays weekly and go live within two to three weeks.
What critics are saying
- BBC reported EMMA failed to understand a stroke patient; Armthorpe surgery decommissioned it by August 2026.
- Healthwatch Rotherham accused EMMA of accent problems on August 19, 2026, inviting NHS scrutiny and complaints.
- If NHS practices judge EMMA inaccessible, procurement freezes spread quickly and the company loses its only market.
What makes QuantumLoopAI unique
- EMMA targets NHS GP surgeries, not generic contact centers, with telephony workflows built for primary care.
- QuantumLoopAI claims DTAC approval, DSPT compliance, and live integrations with eConsult, Accurx, ANIMA, System Connect, Rapid Health.
- The company says it answers hundreds of calls simultaneously and supports 17 languages across NHS settings.
Help us improve and share your feedback! Did you find this helpful?
Benefits
- Remote Work Options
- Stock Options
- Conference Attendance Budget
- Professional Development Budget