Pachyderm has been acquired by Hewlett Packard Enterprise - Learn more

Automate Data Transformations

Data-driven pipelines for machine learning healthcare automotive services unstructured data any data & language

Request a Demo
Pachyderm
Liveperson
Fraunhofer
LogMeIn
Woven Planet
RTL

Automate Complex Pipelines
with sophisticated data
transformations

Kubernetes
Pipeline V2

Orginal Data Sets

V2
  • photos
  • photos
  • photos
  • photos
  • photos
  • photos
  • photos
  • photos
Listening for changes...
Object Store

Transformation Code

V1

  import cv2
  import numpy as np
  from matplotlib import pyplot as plt
  import os

  # edges.py reads an image and outputs transformed image
  def make_edges(image):
    img = cv2.imread(image)
    tail = os.path.split(image)[1]
    edges = cv2.Canny(img,100,200)
    plt.imsave(os.path.join("/pfs/out", os.path.splitext(tail)[0]+'.png?as=webp'), edges, cmap = 'gray')

  # walk images directory and call make_edges on every file found
  for dirpath, dirs, files in os.walk("/pfs/images"):
    for file in files:
        make_edges(os.path.join(dirpath, file))

New Data Sets

V2
  • photos
  • photos
  • photos
  • photos
  • photos
  • photos
  • photos
  • photos
Listening for changes...
Object Store

Automatic Detection

Data-driven pipelines automatically trigger based on detecting data changes.

Version Control

Automatic immutable data lineage and data versioning of all data types.

Autoscaling

Autoscaling and parallel processing built on Kubernetes for resource orchestration.

Automatic Deduplication

Uses standard object stores for data storage with automatic deduplication.

Cloud & On-prem

Runs across all major cloud providers and on-premises installations.

CSV Data Video Data Table Data Genome Data Image Data

Structured & Unstructured

Transform Any Data

Automatic and intelligent versioning of even the largest data sets of unstructured and structured data.

Git-like structure enables effective team collaboration. Full versioning for metadata including all analysis, parameters, artifacts, models, and intermediate results.

Automatically produces an immutable record for all activities and assets.

Transform your Company

Pachyderm is used across a variety of industries and use cases. Pachyderm provides a powerful solution to optimize data processing, MLOps, and ML Lifecycles.

Healthcare & Life Sciences

  • Patient Records
  • X-Rays
  • Genomics
Transform data
  • Modernized Care
  • Improved Diagnostics
  • Genomic Sequencing
  • Optimized Treatment

The scientific method's reproducibility gave birth to the scientific revolution. As biotech firms turn to AI/ML to drive next generation drug discovery, they need a new kind of reproducibility: Pachyderm's data lineage delivers the scientific method data scientists need to create scalable, repeatable experiments now. Read more...

Financial Services

  • Call Recordings
  • Financial Transactions
  • Security Logs
Transform data
  • Fraud Detection
  • Customer Sentiment Analysis
  • Automated Trading
  • Risk Assessment

Machine Learning has made a large impact in Financial Services by addressing a wide range of applications from fraud detection to improved customer service and robo investing tools. Pachyderm’s immutalbe data lineage ensure that all processes can be reproduced reducing risk and enforcing compliance. Read more...

Natural Language Processing (NLP)

  • Documents
  • Text Files
  • Recordings
Transform data
  • Sentiment Analysis
  • Automatic classification
  • Actionable insights
  • Automatic redaction

Detect customer sentiment and analyze customer interactions. Provide insight into efficient customer escalation handling and analyze key words and phases. With Pachyderm's complete pipelining and lineage the model can be refined until true actionable insights are reproduced. Read more...

Video and Image Processing

  • Videos
  • Images
  • Metadata
Transform data
  • Object Detection
  • Motion tracking
  • Feature recognition
  • Color correction

Video and imaging ETL is characterized by large unstructured data sets that can create bottlenecks for teams as they look to productionize and scale. Pachyderm provides and autoscaling deduplication platform that effortlessly scales and deduplicates large datasets saving costs. Read more...

Pachyderm Jupyter Label Studio Spark Tensorflow WB

Data Science & ML Ops

Integrates into your workflow

We understand that you support Data Scientists, MLOps and other infrastructure teams. They will love Pachyderm too!

Data Science Support: Let Pachyderm be the single source of truth for your data. Use familiar Jupyter notebooks to experiment and iterate with your data collaboratively, while always remaining in sync.

MLOps Support: We work with the standard Kubernetes tools, integrate into existing systems and run across all cloud and on-premises providers.

Use Any Language, Transform Any Data

Built for Data Engineers

Pachyderm is container-native, running with standard containerized tooling and allows engineers complete autonomy to use whatever languages or libraries are best for the job.

Pachyderm is data-agnostic, supporting both unstructured data such as videos and images as well as tabular data from data warehouses.

Explore the Docs

import cv2
import numpy as np
from matplotlib import pyplot as plt
import os

# edges.py reads an image and outputs transformed image
def make_edges(image):
  img = cv2.imread(image)
  tail = os.path.split(image)[1]
  edges = cv2.Canny(img,100,200)
  plt.imsave(os.path.join("/pfs/out", os.path.splitext(tail)[0]+'.png'), edges, cmap = 'gray')

# walk images and call make_edges on every file found
for dirpath, dirs, files in os.walk("/pfs/images"):
  for file in files:
      make_edges(os.path.join(dirpath, file))

var cv2 = require('cv2');
var np = require('numpy');
from matplotlib var plt = require('pyplot');
var os = require('os');

// make_edges reads an image from /pfs/images and outputs the result of running
// edge detection on that image to /pfs/out. Note that /pfs/images and
// /pfs/out are special directories that Pachyderm injects into the container.
function make_edges(image) {
  img = cv2.imread(image);
  tail = os.path.split(image)[1];
  edges = cv2.Canny(img,100,200);
  plt.imsave(os.path.join('/pfs/out', os.path.splitext(tail)[0]+'.png'), edges, cmap = 'gray');
}
// walk /pfs/images and call make_edges on every file found
for (dirpath, dirs, files in os.walk('/pfs/images')) {
  for (file in files) {
      make_edges(os.path.join(dirpath, file));
  }
}

package it.polito.elite.teaching.cv;

/**
 * @author luigi.derussis@polito.it Luigi De Russis
 */
public class HelloCV
{
	public static void main(String[] args)
	{
		// load the OpenCV native library
		System.loadLibrary(Core.NATIVE_LIBRARY_NAME);
		
		// create and print on screen a 3x3 identity matrix
		System.out.println("Create a 3x3 identity matrix...");
		Mat mat = Mat.eye(3, 3, CvType.CV_8UC1);
		System.out.println("mat = " + mat.dump());
		
		// prepare to convert a RGB image in gray scale
		String location = "resources/Poli.jpg";
		System.out.print("Convert the image at " + location + " in gray scale... ");
		// get the jpeg image from the internal resource folder
		Mat image = Imgcodecs.imread(location);
		// convert the image in gray scale
		Imgproc.cvtColor(image, image, Imgproc.COLOR_BGR2GRAY);
		// write the new image on disk
		Imgcodecs.imwrite("resources/Poli-gray.jpg", image);
		System.out.println("Done!");
	}
}

import scala.io.Source
import java.io.PrintWriter
​
object RemoveNan {
  def main(args: Array[String]): Unit = {
    // Read the csv file and store it as a list of strings
    val lines = Source.fromFile("/pfs/input_repo/input.csv").getLines.toList
​
    // Filter out rows containing Nan values
    val cleanRows = lines.filter(line => !line.contains("Nan"))
​
    // Write the filtered rows to a new csv file
    val writer = new PrintWriter("/pfs/out/output.csv")
    cleanRows.foreach(line => writer.write(line + "\n"))
    writer.close()
  }
}

Choose Any Data Type

Choose Any Data Type

Integrate with your favorite tools

Integrations

automate your data pipeline

Pachyderm

Automate Pipelines Easily

Easy as 1-2-3

Pachyderm is data-agnostic, supporting both unstructured data such as videos and images as well as tabular data from data warehouses.

Pachyderm is container-native, running with standard containerized tooling and allows engineers complete autonomy to use whatever languages or libraries are best for the job.

Pipelines are intelligently triggered by detecting changes to data, which is all automatically version controlled by the platform.

Any Scale

Process petabytes of data, thousands of jobs, hundreds of models.

Data Pipeline

Transform your data pipeline

Learn how companies around the world are using Pachyderm to automate complex pipelines at scale.

Request a Demo

Transform your Company

Learn how companies around the world are using Pachyderm to automate complex pipelines at scale.