Python, but big.

It's time to ditch your VPN

Churn through a ton of data, no cloud expertise needed

Dask Dataframes for ETL

Dask is faster than Spark. Less painful too.
  • Pandas but massive
  • Easy to set up and deploy
  • Scales comfortably from 10 GiB to 100 TiB
  • Production and Enterprise-ready
Parallelism
GPUs
import coiled
import dask.dataframe as dd

cluster = coiled.Cluster(
    n_workers=20, region="us-east-2"
)
client = cluster.get_client()

# How much did NYC pay Uber?
df = dd.read_parquet("s3://coiled-data/uber/")
df.base_passenger_fare.sum().compute()

# And how much did Uber drivers make?
df.driver_pay.sum().compute()

GPUs, but easy to get

Wish your macbook had 100 NVIDIA GPUs?
  • Access any GPU on any Cloud
  • Super easy to use from your laptop
  • Automatically copies your code and libraries
Parallelism
GPUs
import coiled, torch 

@coiled.function(
    vm_type="g5.xlarge",
    region="us-west-1",
)
def train(trial: int) -> pytorch.Model:
    torch.device("cuda:0")
    
    model = ...

    return model.to_device("cpu")

model = train(123)

Planet Scale Data?
No problem.

  • Xarray clusters in any region
  • Process terabytes of satellite imagery
  • Easy controls minimize and constrain costs
Parallelism
GPUs
import xarray as xr
import coiled
  
cluster = coiled.Cluster(
    n_workers=500,
    region="us-west-2",
    worker_memory="64 GiB",
    spot_policy="spot",
)
client = cluster.get_client()

ds = xr.open_mfdataset(
    "s3://mybucket/all-data.zarr",
    engine="zarr", parallel=True,
)

Any Python Code

Sometimes you just want big parallel for loops
  • Easy to use serverless functions
  • No restrictions on size, region, or account
  • Automatically scales up and cheap as dirt
Parallelism
GPUs
import coiled
import s3fs

@coiled.function(region="us-west-2", worker_memory="128 GiB")
def process(filename: str) -> float:
    with s3.open(filename, mode="rb") as f:
        data = f.read()
        
    output = transform(data)
    
    with s3.open(filename + ".out", mode="wb") as f:
        f.write(output)

filenames = s3.ls("s3://mybucket/*.png")
process.map(filenames)

Coiled runs your Python code on lots of VMs

Code locally, run at scale

Coiled creates ephemeral VMs that match your local environment exactly, spinning up clones in about a minute, copying ...

  • Your code
  • Your software libraries
  • Your working files
  • Your cloud credentials (securely)
Deploy with
Hitech
Deploy with
Luminous
Deploy with
Automation
Deploy with
Glossy

Powerful architecture for easy development

Designed for Python devs, not cloud wizards

Run securely in your cloud

Coiled runs in your cloud account.

Your data stays private and secure.

Run from anywhere

Questions:

  • How does this integrate with notebooks?
  • How do I run production jobs on a schedule?
  • Where do I edit my code?

Answer:

  • Exactly the way you did before.

Coiled runs from where you run Python today.
Integrations are easy.  Just import coiled

Run on any hardware

Use any hardware in any region.

  • Grab GPUs from San Francisco or Frankfurt, wherever has more
  • Want a Terabyte of memory? No problem.
  • Boost cost efficiency with ARM-Graviton and Spot
@coiled.function(
    vm_type="m7i.32xlarge", # 128 Core VMs
    region="eu-west-2",     # in London
    spot_policy="spot",     # using spot
)
def train(...):
    ...

Trusted by data teams large and small.

Delightful to Use

Just ask these beautiful people...

Coiled changed how we work
“The speed is nice, sure, but the real benefit is taking a multi-day effort and finishing it in an afternoon. Coiled changed the character of our work.“
Matt Plough - Software Engineer

Matt Plough

Software Engineer, KoBold Metals

Burst to the cloud

“Quite literally ‘burst to the cloud from your laptop’, everything I've been dreaming of since grad school.“
Eric Ma, Principal Data Scientist

Eric Ma

Data Scientist, Moderna

Laptop to cloud
“Coiled is a game changer for us. It's great to be able to run the same process on your laptop and on the cloud.“

Basile Goussar

Co-founder, Netcarbon

Hard to break
“Coiled is a super simple way to launch a cluster with a simple API that's hard to break.“
Eric Jeske - Head of Data Science

Eric Jeske

Head of Data Science, Telemetry

I never worry about it

“I’ve been incredibly impressed with Coiled; it’s quite literally the only piece of our entire ETL architecture that I never have to worry about.“

Bobby George

Co-founder, Kestrel

Dead simple

“Dask is simple. Coiled made it simpler. Set up is half the battle with the cloud and Coiled made this easy.“

Kenneth Nguyen

CTO, Tasq

The Heroku of data

“Coiled is the Heroku of Data.  Setup was a piece of cake.“

Tim Cull

Leadership Swiss Army Knife, Urban Footprint

Up and running in no time

“My team has started using Coiled this week. Got us up and running with Dask clusters for ad hoc distributed workloads in no time.“

Mike Bell

Data Scientist, Titan

Amazing support

“Coiled support is amazing. I’ll run into an issue and before I have a chance to mention it I have an email in my inbox. You don’t get this kind of support with large companies.“

Katya Potapov

Software Engineer, Floodbase

Hours instead of days

“On my computer this takes days. Now it takes an hour. I had no experience with distributed systems.“

Mohamed Akbarally

Data Scientist, With Marmalade

It just works
“We've been using Coiled in our backend for months and never think about it. It just works.“

Luiz Augusto Alvim

Acoustic Engineer, RPG Acoustical

Fun to use

“Dask and Coiled are natural and fun to use. They're Pythonic.“

Lucas Gabriel Balista

Data Science Lead, Online Applications

Amazing support
“Coiled support is amazing. I’ll run into an issue and before I have a chance to mention it I have an email in my inbox. You don’t get this kind of support with large companies.“
Burst to the cloud
“Quite literally ‘burst to the cloud from your laptop’, everything I've been dreaming of since grad school.“
Eric Ma, Principal Data Scientist

Eric Ma

Data Scientist, Moderna

Dead simple
“Dask is simple. Coiled made it simpler. Set up is half the battle with the cloud...Coiled made this easy.“

Kenneth Nguyen

CTO, Tasq

Hours instead of days
“On my computer this takes days. Now it takes an hour. I had no experience with distributed systems“

Mohamed Akbarally

Data Scientist, With Marmalade

Never worry about it
“I’ve been incredibly impressed with Coiled; it’s quite literally the only piece of our entire ETL architecture that I never have to worry about.“
Florian Jetter - Software Engineer Team Lead

Bobby George

Co-founder, Kestral

Heroku of data
“Coiled is the Heroku of Data.  Setup was a piece of cake.“
Florian Jetter - Software Engineer Team Lead

Tim Cull

Leadership Swiss Army Knife, Urban Footprint

Fun to use
“Dask and Coiled are natural and fun to use. They're Pythonic.“
Florian Jetter - Software Engineer Team Lead

Lucas Gabriel Balista

Data Science Lead, Online Applications

Up and running in no time
“My team has started using Coiled this week. Got us up and running with Dask clusters for ad hoc distributed workloads in no time.“
Florian Jetter - Software Engineer Team Lead

Mike Bell

Data Scientist, Company

It just works
“We've been using Coiled in our backend for months now and never think about it. It just works.“
Florian Jetter - Software Engineer Team Lead

Luiz Augusto Alvim

Acoustic Engineer, RPG Acoustical

Laptop to cloud
“Coiled is a game changer for us. It's great to be able to run the same process on your laptop and on the cloud.“

Basile Goussar

Co-founder, Netcarbon

See Coiled in Action

Use cases, demos, and tutorials to help you get started.

Easy to Get Started

Setup takes about three minutes

To start all you need are cloud credentials and a command line.

Start work on the generous free tier with 10,000 CPU-hours per month, more than enough for most individuals.*
* You still have to pay your AWS, GCP, or Azure cloud costs.
$ pip install coiled

$ coiled login

$ coiled setup aws

Grant cloud access? (Y/n): Y
... Configuring ...
🎉 You're ready to go.

Congrats! You made it this far.

Try Coiled for yourself or get more details.