Berkeley Lab · 11 hours ago
Software Developer (KBase)
San Francisco Bay Area, CA
Full-time
Hybrid
Senior Level
$117K/yr - $146K/yr
5+ years expLawrence Berkeley National Laboratory is looking for a Software Developer to join the US Department of Energy’s Systems Biology Knowledgebase (KBase) team. In this role, you will design, develop, and maintain software infrastructure for AI-driven scientific workflows, contributing to the integration of diverse biological datasets and tools to support large-scale research in microbial ecology and systems biology.
Research
No H1B
Responsibilities
Lead the design and development of AI agents integrated with the KBase Narrative Interface, KBase Apps, and data services
Develop agent-based tools using Python, LangChain, LangGraph, CrewAI, and modern large language models (e.g., OpenAI GPT-5.2, Claude Opus 4.5, DeepSeek, Kimi)
Implement AI agents that interact with the KBase Datastore, a Delta Lake-based lakehouse built on PySpark
Design and maintain agent state management systems, prompt strategies, and benchmarking tools to ensure reliability and performance
Write robust, efficient, and well-tested code to support agent orchestration and pipeline integration
Collaborate with the AI/ML team to advance foundational models and tools for microbial genomics research
Contribute to front-end and back-end integration as needed to expose AI agent capabilities to end users
Develop, test, and maintain documentation and software artifacts in compliance with team standards and quality assurance procedures
Provide technical guidance and mentorship in AI tool design and system integration
Qualification
Required
A Bachelor's Degree (or equivalent knowledge/training) in Computer Science, Engineering, Computational Biology, or a related field and a minimum of 5 years of relevant work experience in AI agent frameworks, data engineering, and software development in service of scientific research or an equivalent combination of education and experience
Experience designing and implementing intelligent agents for scientific or technical workflows
Demonstrated proficiency in Python and AI agent frameworks (LangChain, LangGraph, and CrewAI)
Experience working with large language models (LLMs) and prompt engineering across multiple providers
Skilled in PySpark and data engineering using lakehouse architectures (Delta Lake)
Familiarity with machine learning libraries (PyTorch, TensorFlow, Scikit-learn)
Knowledge of open-source collaboration, GitHub workflows, and Agile practices
Strong grasp of AI agent state management, evaluation, and benchmarking
Familiarity with deploying data/compute-intensive workflows on HPC or cloud research infrastructure
Excellent oral and written communication skills including experience organizing and presenting information to technical and non-technical audiences
Strong analytical skills including experience identifying and solving complex technical problems
Demonstrated interpersonal skills including experience collaborating with a variety of scientific, operations, and technical teams
Preferred
A Master's Degree (or equivalent knowledge/training) in Computer Science, Engineering, Computational Biology, or a related field
Experience fine-tuning large language models or developing custom language models
Experience with scientific reproducibility practices and metadata standards
Familiarity with scientific data analysis and visualization techniques
Benefits
Exceptional health and retirement benefits, including pension or 401K-style plans.
A culture where you’ll belong - we are invested in our teams!
Enjoy a Winter Holiday shutdown every year.
Parental bonding leave (for both mothers and fathers).
Pet insurance.
Company
Berkeley Lab
Berkeley Lab is a national laboratory that creates advanced new tools for scientific discovery.
1001-5000 employees
Funding
Current Stage
Late StageLeadership Team
Mary Barnum, MBA
Business Manager, COO Office
Rebecca Rishell
Deputy Chief Operating Officer
Recent News
MIT Climate Portal - Massachusetts
2025-07-18
Help Net Security
2025-04-15
Company data provided by crunchbase