Large-scale data science for a small team
What do you like best about the product?
I left a large company to join a small one. I was worried that I would not be able to create quality research without the large data infrastructure of the big firm. Saturn provides a scalable data science infrastructure. Their environment is better than what I had access to before. We can spin up clusters and resources at the click of a button. It also integrates with prefect which I am excited to try in the future.
What do you dislike about the product?
Some features are missing if you want to use this for production. With that said it's a matter of time before they roll these out. Things such as shared resources, for example.
What problems is the product solving and how is that benefiting you?
We don't need a dev-ops team to do our research. Our small team has the data science infrastructure that would previously only be available at a large company.
Recommendations to others considering the product:
The saturn team is responsive. Use the saturn slack channel for help.
Powerful yet simple
What do you like best about the product?
Speed to get things up and running, and being able to deploy without help from an engineering team.
The interface is easy to use, and the customer support is great; I had one issue connecting via SSH to my working directory and their CTO took me through how to sort it, step-by-step.
What do you dislike about the product?
The team / organizational functionality outside of the enterprise edition could probably be better.
What problems is the product solving and how is that benefiting you?
Setting up jobs for automating business tasks with python scripts which would be difficult to do on a traditional cloud platform. Running data science models on scalable instances with reliable VM images.
Great platform for parallelizing hyperparameter tuning and k-fold cross validation tasks using Dask!
What do you like best about the product?
Prebuilt tools to spin up clusters easily
What do you dislike about the product?
It wasn't very easy to edit docker image
What problems is the product solving and how is that benefiting you?
Hyperparameter tuning and k-fold cross validation with lightgbm and Optuna for training GBDT models
Recommendations to others considering the product:
Great development environment for data science teams in startups!
Saturn cloud review
What do you like best about the product?
Easy log-in and stable interface. Excellent customer support.
What do you dislike about the product?
No R-Studio interface. It is challenging to run docker inside Saturn.
What problems is the product solving and how is that benefiting you?
Analysis code development. The interface is very stable. The developing environment is easy to be maintained and updated.
On top of that, the support is excellent and can solve any problems!
Scalable notebooks in my own AWS environment
What do you like best about the product?
Saturn makes it easy for our analysts to spin up Jupyter servers of any type on demand. The flexibility in being able to define custom docker images, startup scripts, etc. is phenomenal. I love the integration with Prefect Cloud, as it lets us seamlessly run our data ingestion pipeline using the same code and in the same environment as our analysts do their work. We run Saturn Enterprise in our own AWS environment, which keeps data moving quickly and meets our regulatory requirements of keeping data within our country. And if I ever run into trouble, Saturn's support always goes above and beyond to help out.
What do you dislike about the product?
With Saturn your end-user interface is the open-source Jupyter Lab ecosystem. Don't get me wrong, it's great. But services like DeepNote provide superior collaboration, SQL integrations, charting without code. I stick with Saturn as the best overall solution, but I am envious of their enduser experience.
What problems is the product solving and how is that benefiting you?
I needed an environment where a diverse team of data analysts could run the same notebooks without worrying about their local environment (Mac, Windows, Linux), and data transfer of files to and from cloud storage.They also needed to be able to access high-CPU machines on occasion. All these benefits were realized with Saturn, and then some.
Excellent data science platform
What do you like best about the product?
Saturn Cloud makes it easy to spin up resources for your experiments or analysis. If you need to perform a large experiment on a lot of data, you can easily spin up a Dask cluster with multiple nodes, or if you need to train a machine learning model, it's easy to start a GPU instance. Onboarding new team members is simplified as everyone is using the same environment.
What do you dislike about the product?
Running hyperparameter searches can be done through Dask, but this does not always feel like the most natural way to perform this task. Having a jobs API that allows you to run arbitrary jobs similar to SLURM would be a helpful addition.
What problems is the product solving and how is that benefiting you?
Saturn cloud has allowed us to scale our experiments faster and simplified our way of working.
Streamlining my Data Science Workflow
What do you like best about the product?
Saturn Cloud makes my work so much easier. When I sit down at the beginning of the day, I just want my environment to work. I want my favorite packages installed and available on demand. I want it to be easy to scale my workspace and have it shut down automatically when I'm done. Saturn Cloud solves all of that. Their customer service is also top-notch.
What do you dislike about the product?
Some parts of Saturn are a little wonky because it's run inside Docker containers. This makes containerized workflows difficult to deploy. I also wish it were easier to set up networked programs like Syncthing or database servers.
What problems is the product solving and how is that benefiting you?
Saturn Cloud makes it easy for anyone to spin up an environment for doing data analysis. It saves time and is less error-prone than being responsible for my own environment.
Seminal Development in Python
Saturn Cloud has opened an entirely new door of opportunities for both the average quantitative developer and the deep learning expert.
At my company, I have been able to explore multiple new avenues for alpha generation that would have been unfeasible prior to the scale and speed introduced by Saturn.
I look forward to seeing the array of new applications developed as Saturn fuses the rapid development speed of Python with the raw power of distributed GPU computing while trivializing DevOps in the process.
Seamless transition from local to cluster thanks to SaturnCloud
I used Saturn Cloud to run an NLP pipeline that I had started building locally (AWS t2.2xlarge) using Python (Jupyter notebooks), Dask, and SciSpacy, but which I was beginning to outgrow. Moving the code to SaturnCloud was quite painless -- all I had to do was to switch out the distributed Dask scheduler with the one provided by SaturnCloud, and re-point to S3 instead of local disk for my data. I would also like to thank the Saturn Cloud engineers, they are very professional and responsive, and without their timely help, my project would have taken much longer than it did. SaturnCloud also offers GPU machines for use with RAPIDS, and offers a Jupyter Lab environment as well. If you are using Dask and need to scale out, Saturn Cloud is a great way to do it without having to invest and set up your own cluster.
Great Product
In my opinion, this is the most best cloud hosted jupyter solution out there. The flexibility of scaling up and down as needed is great, as well as the seamless Dask integration. Not to mention the very responsive support team!
