STARLIMS provides leading Laboratory Information Management System solutions that have served customers around the world for nearly 40 years. The role focuses on building AI capabilities into STARLIMS, developing reliable production-grade systems that integrate with real workflows and domain-specific data.
Responsibilities:
- Design and optimize RAG pipelines over domain-specific content
- Improve retrieval quality, ranking, and grounding to reduce hallucinations
- Build evaluation frameworks to measure accuracy and consistency
- Work with fragmented enterprise data and make it usable
- Build AI-powered features integrated into STARLIMS workflows
- Design for reliability (latency, scale, model variability)
- Implement guardrails, fallbacks, and observability
- Manage prompt evolution, model drift, and regressions
- Develop systems that can reason, call tools, and execute multi-step tasks
- Integrate with internal APIs, developer tooling, and external systems
- Design controlled execution paths for automated actions
- Build and evolve a VS Code-based development environment replacing a legacy desktop IDE
- Support development across a custom application framework and domain-specific language
- Develop AI-assisted workflows for generating, understanding, and debugging applications
- Integrate LLMs with platform APIs, developer tools, and contextual knowledge systems
- Design extensible mechanisms for AI-assisted multi-step development tasks
- Build and operate backend services on AWS (Lambda, API Gateway, DynamoDB, etc.)
- Own system architecture and key technical decisions
- Contribute to infrastructure-as-code and deployment pipelines
Requirements:
- 6+ years of software engineering experience, including production systems
- Experience building RAG systems beyond prototypes
- Strong understanding of LLM behavior, limitations, and failure modes
- Experience with LLM APIs and prompt/system design
- Solid backend and cloud experience (AWS or equivalent)
- Proficiency in TypeScript and/or Python
- AI evaluation pipelines and metrics
- Agent/tool-use systems or similar architectures
- Developer tooling (VS Code extensions, language tooling)
- Workflow automation platforms (n8n, Zapier, etc.)
- Experience in regulated or domain-heavy systems
- Scaling vector search systems
- Experience with containerization and orchestration (ECS, EKS, Kubernetes)
- Infrastructure as Code (Terraform or similar)