Private & Confidential
Seqera × Anthropic

The Workflow Engine
for AI in Life Sciences

Nextflow is the most widely adopted workflow engine in life sciences. Claude is the most capable AI for scientific reasoning. Together, it closes the loop on AI-assisted discovery and execution.

March 12, 2026
AstraZeneca GSK Pfizer Roche AbbVie Moderna

Daily workflow runs have grown from ~20K in early 2024 to exceeding 200K

Daily runs 14-day avg AI era
Source: Nextflow telemetry
5,000+
Open Source contributors
2.5M+
Monthly runs
130+
Life Sciences customers
13/20
Top Biopharma
What exists today
Anthropic's Nextflow Skill

Anthropic recently launched the Nextflow Development Skill as part of the Claude for Life Sciences initiative. It enables bench scientists to run omics analyses without specialized bioinformatics training, but is currently limited to local execution of a subset of nf-core pipelines. The next step is agentic orchestration at enterprise scale, on any pipeline and with full traceability.

BioPipelineBench assesses AI's ability to execute bioinformatics workflows spanning targeted and long-read sequence analysis, metagenome assembly, and chromatin profiling. The rate of improvement signals practical viability.

Source: Anthropic BioPipelineBench
The next step
Enterprise-scale orchestration via Seqera Platform — any pipeline, any cloud, full traceability.
Vision
AI-Native Scientific Computing
Intelligence Layer (Claude)
Reasoning · Design · Interpretation
Execution Layer (Nextflow)
Orchestration · Scale · Pipelines
Infrastructure Layer (Seqera)
Multi-cloud · Audit · Governance
Momentum
Transforming Bioinformatics with Claude

Seqera is now rewriting the fundamental pipelines for biology. With Claude Opus 4.6, we developed RustQC, a quality control pipeline that was rebuilt from the ground up. We were able to achieve a remarkable 21x performance improvement.

RustQC Benchmark Results
Agents
Agent-driven pipeline optimization

Seqera is using AI agents to massively parallelize workflow optimizations.

Systematic exploration
AI agents evaluate thousands of parameter combinations across tools, references, and thresholds to find optimal configurations for each dataset.
Parallel execution
Nextflow's native parallelism combined with Claude's reasoning enables simultaneous optimization across pipeline stages at enterprise scale.
Quality-driven iteration
Agents compare outputs against gold-standard references, iterating on parameters until results meet or exceed expert-level benchmarks.
BioML training data
Optimized pipeline outputs become high-quality training data for biological ML models, creating a flywheel between workflow execution and AI improvement.
Execution Environment
Seqera as a RF environment

~1B bioinformatics task executions on Seqera Platform

  • Rich reward signal — every task has a clear success/failure outcome, resource allocation, and performance metrics
  • Enormous action space — tool selection, parameter configuration, resource allocation, error handling
  • Closed-loop feedback — agent proposes → platform executes → outcome observed → agent improves
  • Domain diversity — covers the full breadth of computational biology
Next steps
Next Steps
1
Extend the Nextflow Skill
Add Seqera Platform as the execution backend — enterprise-scale, any cloud, full governance
2
Run BioPipelineBench
Seqera's Claude-powered agent against the benchmark — demonstrate the integration
3
RL Data Partnership
Structured access to Seqera's execution metadata for Claude model improvement
4
Developer Co-Marketing
Nextflow Summit, nf-core community, joint content
Seqera x Anthropic

By joining Seqera's expertise in Life Sciences with Anthropic's scientific reasoning, we can radically change the trajectory of human health.

"Biology is probably the area where scientific progress has the greatest potential to directly and unambiguously improve the quality of human life."— Dario