Today we’re excited to be unveiling something totally new in the world of data science and AI development. We’ve been building in stealth for the last two years, and we're finally opening the door to let the world in.

Our Mission

Our mission at Zerve is to Elevate the impact of Data Scientists. Why? A few reasons.

1. My two co-founders Greg & Jason, two PhD level statisticians, are two of the smartest people I know. They have tales of not being able to get their organization's AI initiatives over the line, or crazy inefficiencies in the process while doing so. They have lived the problems. Teams separately working on local notebooks and meeting twice a day to share findings, and emailing/zipping the latest data to other team members. Inefficient and not fit for purpose in a world that will demand impact from Data Science teams like never before.
2. We’ve spent 100s of hours interviewing “Code-First Data Users” who not only confirmed the efficiency losses we knew, but shed light on even more. We heard all too often that the Data Scientist's impact was set up to be distinctly unimpactful—a screenshot of a Matplotlib chart in a deck, an uninspiring data app, or a handover that resulted in a complete re-coding of the original work.
3. We believe to truly unlock the opportunities in AI, LLMs, and Data Science over the next decade, expert Data Scientists will be the difference makers, and we set out to build a platform that puts Data Scientists at the heart of the innovation.

The Trend Over the Last Decade

The trend over the last decade has been moving toward low-code/no-code solutions for developing data science. Software vendors have developed a variety of tools that supposedly allow anyone to train sophisticated models and interact with data without writing code. It has become obvious to anyone who has actually deployed a data science project that these tools are inflexible and unsuitable for serious work. That’s because virtually all data science projects require code, and that’s why Zerve is built for code-first data teams. LLMs can help build data solutions further, but they work best in the hands of code-first users who know the domain and the stack.

The coding toolkit for data, though, is problematic. Data engineers might be working in Snowflake or dbt, writing SQL. Data scientists typically start in a notebook like Jupyter. The problem with notebooks is that they are brittle. Jupyter was developed by academics to be used as a classroom scratchpad. Jupyter is wonderful for exploration, but it’s undeployable—meaning that if you’re working on something serious, you’ll eventually need to move to an IDE like PyCharm or VS Code.

Because of the architecture of notebooks, it’s easy to get them into a bad state. In the example below, by running the cells out of order, you can end up with incorrect results. That’s why every modern notebook has a “restart kernel” button: global state makes production work infeasible and risky.

Since notebooks can’t handle production workloads, developers turn to scripting environments or IDEs. These tools are terrific for writing stable code, but they were built for software developers and are less suited to data exploration. Maybe that’s why the Jupyter plugin for VS Code has been downloaded over 40 million times.

Given AI’s importance at the board level, it’s surprising that this fragmented toolkit is involved in almost every serious data science project.

That’s where Zerve comes in. Zerve is a Data Science Development Environment built on a novel architecture, giving you the benefits of interactive analysis without sacrificing the stability of a scripting environment. With Zerve you can explore, build, and deploy in the same environment. To truly do this, the underlying architecture had to ensure stability. Let's explore that.

The New Zerve Architecture

The foundation of Zerve’s architecture is separating storage from compute. In Zerve, code is organized as a DAG, and executes left to right. Each block inherits the state of preceding blocks. When you run code, Zerve launches serverless compute. The block’s state after execution is cached, serialized, and stored on disk so it’s ready for the next block. This produces important benefits.

Benefit: Stability

There is no restart kernel button in Zerve. You can’t get a canvas into a bad state, so there’s no need to restart. Cloud notebooks improved collaboration but didn’t fix global state; multiple users can make state drift worse. Zerve is guaranteed to produce a consistent result no matter how code is executed, single or multiplayer.

Benefit: True collaboration

In Zerve, collaboration is real time and stable. Run blocks concurrently, work in multiple languages, leave comments and mentions, and soon integrate seamlessly with Git-based workflows for version control and change management.

Benefit: Language interoperability

Because Zerve serializes outputs, different languages can use the same basic data types in the same project. For example, data frames are serialized as Parquet, so SQL, R, and Python interact with the same artifacts without fragile code translation. In Zerve, data/ML engineers, R users, and Python users can develop side by side.

Benefit: Seamless multi-processing

Zerve’s serverless tech lets you run as many blocks in parallel as needed. Compute is provisioned on demand and torn down when idle, so you save both time and cost—no servers to babysit.

Benefit: Persistent artifacts

Zerve serializes models, datasets, and other artifacts by default. They’re accessible across your project and from external applications, simplifying the path from development to production.

Start building in Zerve today

Zerve is free to use, with generous storage and compute on the free plan. Whether you’re querying with SQL, visualizing with R, or deploying models in Python, Zerve enhances your workflow, boosts productivity, and enables real collaboration across your data projects.