- Company Name
- Homeprotect Home Insurance
- Job Title
- AI Research Engineer
- Job Description
-
Job Title: AI Research Engineer
Role Summary:
Design, build, and deploy AI-first products and agents on Google Cloud Platform to enhance internal productivity and automate claims processing for a non‑standard home insurance provider.
Expactations:
- Deliver measurable business value through AI enhancements.
- Collaborate with cross‑functional stakeholders, especially Claims Operations.
- Continuously evaluate and improve AI behaviour and performance.
- Maintain high standards of code quality, observability, and cost management.
Key Responsibilities:
- Develop AI agents (e.g., Alfred, claims automation) using GCP ADK, Vertex AI, BigQuery, and CloudRun.
- Build robust Python backend services and integrate REST/JSON APIs.
- Implement LLMs and embeddings (OpenAI, Gemini, Llama, Claude), prompt design, function calling, RAG patterns, and vector databases.
- Design and refine agent evaluation frameworks: completeness, accuracy, client outcome metrics, A/B testing, and cost/latency monitoring.
- Work with Claims Operations to showcase AI functionality and impact.
- Apply DevOps best practices: Git workflows, CI/CD (Azure DevOps), Docker, observability (logging, metrics, tracing).
Required Skills:
- Strong Python programming and backend service development.
- Hands‑on experience with large language models, embedding techniques, prompt engineering, RAG, and vector databases.
- Proficiency in GCP services: Vertex AI, Model Garden, BigQuery, CloudRun Functions.
- Document and data handling: text normalization, chunking, metadata/ACL, Postgres, Redis, Azure Blobs.
- DevOps knowledge: Git, CI/CD pipelines, Docker, observability, evaluation frameworks, A/B testing, and performance monitoring.
- Effective communication and stakeholder collaboration skills.
Required Education & Certifications:
- Bachelor’s degree or higher in Computer Science, Engineering, Data Science, or a related field.
- Certifications in Google Cloud Platform (GAAP, Vertex AI) or AI/machine learning are desirable but not mandatory.
New malden, United kingdom
Hybrid
Mid level
10-02-2026