MeridianLink
AI Engineer II ( AI Platform)
About this role
Mid-level AI engineer responsible for designing and maintaining platform components that enable AI-powered features across MeridianLink's products. You'll build model serving layers, RAG infrastructure, prompt management systems, and evaluation pipelines while collaborating with product teams and establishing integration standards.
What you'll do
- Implement AI platform components including model serving, retrieval infrastructure, and prompt management systems
- Develop and maintain APIs for product teams to integrate AI capabilities into applications
- Build observability and monitoring services to track output quality, latency, cost, and system health
- Collaborate with product engineering teams to translate AI integration requirements into platform features
- Contribute to documentation, runbooks, and reference implementations for platform adoption
- Implement evaluation pipelines and safety guardrails appropriate for a regulated financial services environment
What they're looking for
- Backend engineering (Python, C#/.NET, Java, or Node.js)
- LLM and ML model integration in production systems
- Retrieval-augmented generation (RAG) and vector databases
- Cloud-managed AI services (AWS Bedrock, Azure OpenAI, etc.)
- API design and event-driven architectures
- System design and data structures
- Platform and shared services architecture
- Monitoring and observability for AI systems
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
MeridianLink
MeridianLink builds a multi-product SaaS platform with foundational services for identity, authentication, and data integration. The company is hiring Software Engineers, AI Software Engineers, and Data Engineers to develop cloud-based solutions, scalable data pipelines, and platform infrastructure.
View all jobs at MeridianLinkLikely interview questions
- Walk us through a production LLM integration you've built — what were the key challenges around latency, cost, or output quality?
- How would you design a retrieval-augmented generation pipeline that serves multiple product teams with different data sources?