🔥 Workflows by Relevance AI

This is a home for all of RelevanceAI's workflows as seen in the dashboard. Sign up and getting started here!

Workflows provide users with a guide on how to run relevant code in Relevance AI. It provides a guided interface through Relevance AI Features.

Some of these features include:

Clustering
Dimensionality Reduction
Labelling/Tagging
Launching Projectors

How to add a workflow

Add a subfolder and move notebook there and push the notebook
Modify scripts/manual_add_to_db.py and add a new document in the DOCS variable and commit/push the script. Make sure to see what the other documents are doing and copy that structure else things may error!
Create PR

🔥 Features

Fast vector search with free dashboard to preview and visualise results
Vector clustering with support for libraries like scikit-learn and easy built-in customisation
Store nested documents with support for multiple vectors and metadata in one object
Multi-vector search with filtering, facets, weighting
Hybrid search with support for weighting keyword matching and vector search ... and more!

🚧 Development

🛠️ Requirements

Python ^3.7.0
AWS CLI v2 - if you need to upload new workflows to prod

Getting Started

To get started with development, install the dev dependencies

❯ make install

Uploading workflows

Make sure your AWS SSO profile and creds configured in ~/.aws/config.

Install yawsso or similar to sync API creds w/ SSO profile needed for CDK. Set the AWS_PROFILE env var to bypass having to specify --profile on every AWS CLI call.

❯ ENVIRONMENT=development AWS_PROFILE=Relevance-AI.WorkflowsAdminAccess make upload

Testing

Then run testing using:

Don't forget to set your test credentials!

export TEST_ACTIVATION_TOKEN=<YOUR_ACTIVATION_TOKEN>

## For testing core workflows
export WORKFLOW_TOKEN_CLUSTER_YOUR_DATA_WITH_RELEVANCE_AI=<DASHBOARD_BASE64_TOKEN_FROM_CLUSTER_WORKFLOW>
export WORKFLOW_TOKEN_VECTORIZE_YOUR_DATA_WITH_RELEVANCE_AI=<DASHBOARD_BASE64_TOKEN_FROM_VECTORIZE_WORKFLOW>
export WORKFLOW_TOKEN_REDUCE_THE_DIMENSIONS_OF_YOUR_DATA_WITH_RELEVANCE_AI=<DASHBOARD_BASE64_TOKEN_FROM_DR_WORKFLOW>
export WORKFLOW_TOKEN_CORE_SUBCLUSTERING<DASHBOARD_BASE64_TOKEN_FROM_SUBCLUSTERING_WORKFLOW>

Run test script

tests all notebooks in workflows
outputs error notebook_error.log

❯ python scripts/test_notebooks.py

## Testing indiv notebook
❯ python scripts/test_notebooks.py --notebooks subclustering/core_subclustering.ipynb

More Helpful Commands

❯ make help
Available rules:

clean               Delete all compiled Python files
install             Install dependencies
lint                Lint using flake8
test                Test dependencies
update              Update dependencies
upload              Upload notebooks to S3 and update ds

RelevanceAI / workflows