BioArk Genome Engineering Operating System (BioArk GEOS)
Project Vision
• Current CRISPR software mainly focuses on guide RNA design or individual bioinformatics
tasks, while successful genome engineering requires engineering strategy selection, feasibility
analysis, laboratory planning, sequencing validation, quality control, and continuous
optimization.
• The BioArk AI CRISPR Engineering Platform is an AI Engineering System implementing a
Design–Build–Test–Learn (DBTL) workflow.
• Bioinformatics is treated as one component of a larger engineering ecosystem.
• The platform leverages biomedical AI systems (e.g. Biomni) for biological reasoning while
BioArk develops proprietary engineering intelligence, laboratory planning, sequencing
interpretation and continuous learning.
AI Systems Integration & Loop Orchestration Engineers
- Role Summary: You will serve as the core orchestrator of the BioArk Genome Engineering Operating System, responsible for bridging all individual agent layers into a unified, continuous Design-Build-Test-Learn (DBTL) loop. Your focus is strictly on system-wide state management, inter-agent communication, and automated workflow orchestration.
-
Key Responsibilities:
- Design and deploy multi-agent orchestration frameworks (utilizing tools like LangGraph, CrewAI, or AutoGen) to manage the state and transition logic between Layer 1 through Layer 8.
- Implement robust “Human-in-the-Loop” (HITL) and “Agent-in-the-Loop” checkpoints, ensuring that critical outputs (e.g., Layer 3 strategy decisions or Layer 4 feasibility flags) can be paused for manual review before proceeding to laboratory execution.
- Develop standardized API contracts, JSON schemas, and data pipelines to guarantee seamless, loss-less handoffs of biological metadata between bioinformatics agents, laboratory planning agents, and continuous learning databases.
- Architect automated feedback loops so that failure diagnoses from Layer 6 (Test) immediately trigger re-evaluations in Layer 3 (Design) and Layer 4 (Build), fueling the continuous learning cycle.
- Implement system-wide observability, error-handling routines, and fallback mechanisms to ensure the platform remains stable if a specific sub-agent fails or hallucinates.
-
Deliverables:
- A fully functional, end-to-end DBTL orchestration pipeline integrating all 8 layers.
- Standardized inter-agent communication protocols and state-management infrastructure.
- System-wide observability and agent telemetry dashboards for tracking workflow progress.
- Required Skills: Advanced multi-agent orchestration (LangGraph, AutoGen, etc.), enterprise system architecture, complex state management, API design, and backend engineering (Python/Go).
- Target Core Agents & Tools: LangGraph, CrewAI, Model Context Protocol (MCP), event-driven architectures (Kafka/Event Grid), and system observability tools (Prometheus/Splunk).
