Lead AI Engineer (Scientific Multi Agent Systems)
Responsibilities
- Design and build multi-agent systems that integrate scientific tools, computational workflows, MCP-compatible services, APIs, and cheminformatics libraries such as RDKit.
- Build containerized deployment pipelines (Docker) with proper observability (Langfuse), logging, and lifecycle management.
- Collaborate on the development of agent evaluation frameworks, including automated testing, benchmarking, and performance monitoring.
- Develop data preparation and engineering pipelines across structured, semi-structured, and unstructured sources, working alongside the data team's stack (Snowflake, Airflow, DBT, PostgreSQL, Oracle).
- Contribute to CI/CD workflows and maintain code quality through code reviews, modular design practices, and technical documentation.
- Follow architecture, security, and engineering patterns established by the team.
Requirements
- Strong Python software engineering.
- Experience building applications based on LLMs, tool-calling, and agent frameworks.
- Experience developing multi-agent or workflow-oriented AI systems.
- Experience with evaluation frameworks, automated testing, and performance benchmarking.
- Docker and containerized application development expertise.
- Comfortable working with databases and data engineering workflows.
- Ability to collaborate in a cross-functional scientific environment.
- Availability to work until at least 1:00 PM EST.
Nice to have
- Langfuse or another AI observability platform experience.
- Snowflake Cortex and/or Amazon Bedrock experience.
- AWS ECS/Fargate deployment experience.
- Data engineering stack experience: Snowflake, Airflow, DBT, Oracle.
- GitLab-based development practices experience.
- Experience in drug discovery / cheminformatics: RDKit, chemical structure analysis, assay data interpretation, scientific data visualization, retrieval over structured and unstructured research sources.
We offer
- Competitive compensation
- Flexible working hours
- Continuous education, mentoring, and professional development programs
- A team with an excellent tech expertise
- Contract through the end of the year, with a possible extension based on project needs and performance.
