Skip to content

Polly Stack

Polly is a cloud based Software as a Service (SaaS) solution. Below details out various layers of Polly.

Data Harmonization & Ingestion

  • Polly Pipelines: Workflow orchestration system(supports Nextflow)

  • Polly Curation: Manual and automated data curation

Data Storage & Management

  • Polly Atlas: For storing structured data in a data model

  • Polly Workspaces: For storing unstructured data

Data Consumption

  • Polly Apps: Custom apps running on a highly scalable infra (Hail, IGV)

  • Polly Notebooks: Data exploration/analysis using Jupyter notebooks

Infrastructure

Polly’s infrastructure is powered entirely by managed cloud providers - AWS. Backed by AWS massive infrastructure, Polly is fast, secure, highly available & scales seamlessly.

It uses managed cloud services like AWS S3, DynamoDB, RDS (running PostgreSQL), AWS ElasticSearch Service, AWS Lambda, Amazon API Gateway, Amazon Cloudfront, Amazon Cognito. Polly’s computational infrastructure is powered by Kubernetes used to orchestrate Docker containers on both AWS. It is highly scalable and compatible with all cloud providers.

Micro-services & Serverless architecture

Various parts of Polly are built and deployed as micro services. This allows for better fault isolation and resilience. A focus on managed serverless components like API Gateway, Lambda functions, S3, DynamoDB, S3 etc also allows Polly to scale seamlessly under high load.

Polly’s micro-services architecture pattern allows the development teams to choose different technologies that are best suited to those parts. Polly has been built ground up using Python3.6, Angular11, NodeJS, R & R-Shiny.

Third Party Integrations

File uploads from Polly front-end is powered by Filestack. Sentry integration allows Polly developers to monitor errors and fix them faster. Mixpanel is used for gathering usage analytics which help product owners refine and improve platform capabilities and user journeys.