75K–85K GBP / year
Software Safety Engineer
QAOn-site — London, UK
Published on 2026-09-23
These details were extracted automatically from the original listing. They may not be complete or fully up to date — it's worth checking the original job post.
About this role
Assess medical-device software lifecycle processes and technical documentation to identify safety risks and evidence gaps. Use AI-assisted workflows to support rigorous certification decisions, and collaborate with product, engineering, and machine-learning teams to improve the tools and expand regulatory coverage.
About the company
Scarlet is a notified body that certifies AI and software medical devices under EU MDR and UK MDR without unnecessary delays.
The team
Devices and Software teams
What you'll do
- Assess customer software lifecycle processes and technical documentation
- Track and act on regulatory research, standards, and guidance
- Work with product, engineering, and applied machine-learning teams to improve systems
- Support expansion into new jurisdictions and certifications
- Create customer-facing content explaining regulatory topics
- Provide regulatory insights to customers and prospects
What we're looking for
- Degree in computer science or a related discipline
- At least 4 years’ experience in medical-device or other safety-critical product manufacturing, auditing, or research
- At least 2 years’ experience developing safety-critical software
- Practical knowledge of software development lifecycle processes, verification and validation testing, configuration management, and cybersecurity
- Exceptional written and verbal communication skills
Nice to have
- Knowledge or training in IEC 62304 and IEC 81001-5-1
- Practical experience with generative AI, such as LLMs, embeddings, RAG, fine-tuning, evaluation, or monitoring
Benefits
- Equity
Hiring process
- 30-minute introductory call
- 30-minute interview with a Software Safety Engineer
- Technical task and 1-hour interview
- Three 20-minute culture and values interviews
- Offer and referencing
