AI Engineer

Stack

Responsibilities

• ▹Build agent capabilities for planning, tool use, memory, and context management, and ship them into production
• ▹Integrate agents with internal and external tools and data sources (retrieval systems, structured datasets, lab/biomed APIs, spreadsheets, search), with robust schemas and safeguards
• ▹Develop quality and evaluation systems, including unit, regression, and scenario/benchmark tests, telemetry, and automated scoring
• ▹Collaborate with scientists to analyze failure modes and improve performance
• ▹Partner with the knowledge and ontology team to ensure outputs are source-traceable and compliant with provenance standards
• ▹Implement safety measures, guardrails, and sandboxed execution for risky operations
• ▹Optimize performance and reliability through profiling, idempotency, retries, rate limiting, and uptime management
• ▹Instrument data pipelines for supervised fine-tuning and reinforcement learning when needed
• ▹Contribute to the agent platform, including services, APIs, orchestration, CI/CD, and observability
• ▹Example first-90-day projects: a multi-tool agent for long-horizon scientific tasks with memory and self-correction; automated citation enforcement with source checking, freshness validation, and provenance display; an evaluation dashboard tracking competency pass rates, latency, and failure modes

Requirements

• ▹Experience building production software in Python and/or TypeScript, with strong systems and API design skills (FastAPI, gRPC, GraphQL, or similar)
• ▹Proven experience shipping LLM applications or agentic systems (tool use/function calling, retrieval/RAG, structured outputs, evaluation, or observability)
• ▹Familiarity with agent/orchestration frameworks (e.g., LangChain, LangGraph, AutoGen, CrewAI, MCP) and vector databases (FAISS, Weaviate, Pinecone)
• ▹Experience with cloud infrastructure and containers (AWS, GCP, or Azure), Docker/Kubernetes/Terraform, CI/CD, and production telemetry
• ▹Ability to translate research prototypes into robust, scalable systems

Nice to have

• ▹Experience with fine-tuning and reinforcement learning (RL, RLAIF, RLHF), including reward design and offline evaluation
• ▹Familiarity with benchmarks such as SWE-Bench, OS-World, or tau-bench
• ▹Knowledge of retrieval and knowledge systems, including schema and ontology design, entity modeling, and provenance tracking
• ▹Background in agentic system safety and security (sandboxing, isolation, permissions, auditability)
• ▹Exposure to life sciences or scientific computing and collaboration with domain experts

Soft skills

What we offer

• ▹Opportunity to shape the foundation of how AI collaborates with human scientists
• ▹Real-world scientific impact
• ▹Weekly quality reviews with scientists
• ▹You are encouraged to apply even if you do not check every box

About the company

Bio is a decentralized science protocol that helps launch and grow AI-driven biotech research. Scientists can raise funds, create value from their work, and distribute that value directly to their communities. Since 2023 it has directed over $50m to researchers.