DEV Community

Cover image for How Pachyderm Parses the Pipeline
Nočnica Mellifera for RudderStack

Posted on

How Pachyderm Parses the Pipeline

cover image by Hp2850

How can we make building ML/AI pipelines easy for all dev teams? That's the problem Pachyderm sets out to solve.

Pachyderm was designed to make building and managing end-to-end ML/AI pipelines easier, regardless of their size and complexity. With Pachyderm, you can track your data lineage and bring together version control for your data with the tools, languages, and frameworks of your choice, to build scalable data science pipelines.

Pachyderm uses product usage data — again collected, routed, and warehoused with RudderStack Event Stream — from within the Pachyderm Hub, its SaaS platform, for product analytics and optimizing the platform. They then leverage their event stream and product usage data along with non-event data from their cloud tools like Salesforce, HubSpot, Zendesk, Slack, and Google Analytics — collected and warehoused with tools like Fivetran and Stitch.

However, a tool with so many neat tricks up its sleeve will always be in danger of flummoxing initial users. Simple actions like setting up a workspace may not be obvious when users are just getting started.

That's where Rudderstack comes in: data pipeline tools from Rudderstack make it easy for the ML experts at Pachyderm to measure and scale their usage data, find accounts in jeopardy because the user hasn't done basic setup, and address only those accounts that need it.

Top comments (0)