/ Union for Bioinformatics

Accelerate Biotech Research & Engineering

Union, a managed version of the workflow orchestrator Flyte™, enables biotech workflows like next-generation sequencing and improves collaboration between researchers and engineers while reducing compute time and lowering overall workflow costs.

Get a demo

“With Flyte™, we want to give the power back to biologists. We want to stand up something that they can play around with different parameters for their models because not every ... parameter is fixed. We want to make sure we are giving them the power to run the analyses.”

— Krishna Yeramsetty, Principal Data Scientist at Infinome (now Inscripta)

What is Flyte™?

Flyte™ is an orchestrator that enforces an architectural design pattern for data and ML workflows. Supported by this pattern, computational biologists, biotech researchers and their engineering teams can collaborate effectively while each team uses its preferred tools and frameworks.

Why Flyte™

Flyte™ meets all the research and engineering requirements for faster, more efficient biotech workflows, like next-generation sequencing faster.

Infinitely scalable

Effortlessly scale biotech workflows by leveraging Union

Built for scale

Union is built on the Kubernetes-native Flyte™ Engine

Versioned workflows

Preserve history of your analytics and roll back if needed

Multi-tenancy

Collaborate with multiple teams on a centralized platform

Dynamic workflows

Researchers gain much needed flexibility through Flyte™’s dynamic workflows

Language agnostic

Researchers and engineers can write workflows in any language

Strong typing

Bulletproof your workflows with strong compile-time checks

Data lineage

Observe data end to end

Parallelism

Inherent parallelism reduces wait times

Maximal efficiency

Avoid rerunning workflows if assumptions haven’t changed

Ergonomic development experience

Union supports a smooth local development experience to test your bioinformatics workflows thoroughly and promote them seamlessly to production. Flyte™’s single CLI command makes the transition from development to production effortless. Its multi-tenant architecture streamlines every aspect of the process, from collaborating with others to separating domains.

Tune compute resources & storage

Biotech engineers recognize that provisioning resources can be a time-consuming task, whether performed directly on the cloud or by managing infrastructure. Flyte™, the orchestrator for Union, allows allocation of resources and storage with a single line of code.

This simplifies the segregation of memory-intensive processes from those that are less resource-intensive, promoting efficient use of resources. By using Union, workflows can be optimized for performance and cost-effectiveness, enabling you to focus on what is most important—your data and analysis.

“Flyte™ has this concept of immutable transformation—it turns out the executions cannot be deleted, and so having immutable transformation is a really nice abstraction for our data-engineering stack.”

— Jeev Balakrishnan, Software Engineer at Freenome

Extensible

Many biotech workflows use file formats and data types that are unique to the industry. Researchers and engineers alike can use Flyte™’s extensible type system to extend, customize and compose bioinformatics-specific types to handle genomic, proteinomic and multi-modal data types.

Optimized

Because Flyte™ is language-agnostic, researchers aren’t bound to the language of the Flytekit Python SDK; they can use the best tools for the heterogeneous workloads required in biotech, whether they’re implemented in bash scripts, C++ libraries, Python or R.

Historical

Biotech researchers rely on historical machine learning runs to revisit and compare analyses. Union’s Flyte™ engine ensures that you can reproduce your experiments through versioning and containerization. With Union, researchers can seamlessly access and re-run previous experiments.

“You can say, ‘Give me imputation’ and [Flyte™ will] launch 40 spot instances that are cheaper than your on-demand instance that you're using for your notebook and return the results back in memory.”

— Calvin Leather, Staff Engineer and Tech Lead at Embark Veterinary