Introducing SigTech platform v2#

Introducing SigTech platform v2, the latest version of the SigTech platform!

Changes in v2#

Changes in v2 of the SigTech platform include:

  • SSO sign-in

  • Upgraded strategy deployment

  • Built-in GitHub integration with Bring Your Own Repo (BYOR)

  • Improved scheduling functionality

  • Time zone handling

  • Clearer job logs

  • New .export() function combining strategy and function jobs

  • Exact mirroring of research and execution environments, including dependencies and file systems

  • Only Framework v9 supported

  • Batch script and pip install no longer required

Improvements over v1#

Platform v2 presents some notable upgrades from v1, such as:

Easier sign-in process via SSO#

Platform v2 now supports single sign-on (SSO) to enable a more streamlined and secure sign-in process. You can use your organization’s credentials to securely authenticate and sign in to the SigTech platform.

Upgraded strategy deployment#

Replacing Production in v1, the Deployments section in Platform v2 offers a host of enhancements to improve the quant workflow and strategy execution process.

  • Direct deployments within the UI: Strategies created in your research environment can be seamlessly executed in our code sandbox. Previously, you had to use Python classes to handle scheduling actions. Now, you can schedule a strategy run directly in the user interface (UI). This enables a much more intuitive user experience.

  • Improved scheduling functionality with time zone handling: As well as manually running a strategy, you can create a custom schedule that automatically runs a strategy in the future. You can also now schedule strategies in different time zones.

  • Clearer strategy visualizations: Results can be displayed in graphical or tabular formats, adding a visual dimension to your analysis.

  • Ability to combine strategy and function jobs: New .export() function combines both strategy and function jobs.

  • API integration: Fetch run outputs and trigger strategy execution via SigTech API call.

  • Mirroring of research and execution environments: The two environments now exactly mirror each other, including dependencies and file systems. This enables a much more seamless transition from strategy ideation to implementation, as your executed code will run exactly the same as it did in research.

Learn more

For a more detailed dive into deploying a strategy in Platform v2, visit the strategy deployment section in our Migration guide.

Improved GitHub integration with Bring Your Own Repo (BYOR)#

Platform v2 introduces built-in GitHub integration, replacing the limited Git functionality of Platform v1. This upgraded solution is called Bring Your Own Repo (BYOR). It lets you store any scripts and notebooks created in the SigTech platform in your own GitHub repositories, as well as use Git-based version control tools in the platform.

This is done via the new SigTech GitHub App, which users will need to install in their GitHub organization. It includes plenty of useful Git features, such as:

  • Git version control

  • Synchronization for v2 workspace to your GitHub remote repo

  • Per-user commits

  • Branching and branch protection rules

  • Fetching changes from your GitHub remote repo

  • Collaboration via pull request creation

Better data browser#

The upgraded data browser makes it easier and quicker to view the market data you have access to.

  • Easier-to-navigate menus: Redesigned menu system enables smoother and easier navigation through the asset class and instrument data available to you.

  • Improved search functionality: You can now search for specific instruments within an asset class, as well as apply filters to narrow down your search.

  • Instrument-level specifications: When retrieving data for a financial instrument, you can choose to specify certain properties relevant to the asset class you’re working with.

    • Futures: Contract year, contract month, interval, data source

    • Equities: Exchange, content, data source

    • ETFs: Content, data source

    • Bonds: Year of issue, bond ticker, interval, data source

    • FX: Interval, data source

    • Indices: Content, data source

    • Macroeconomic: Indicator, data source

Clearer debugging of jobs#

You will also be able to view the full log even if the job fails, allowing for easier debugging.

Moving from v1 to v2#

v1 users will have the option to move to v2. For more information about migrating between platform versions, read our Migration guide.

Updated strategy deployment process#

Due to improvements made in strategy deployment, the way you run a strategy and view its output has changed. The process has been streamlined to enable users to schedule a run directly within the interface.

Note: All GitHub org members are able to view everything listed in the organization’s SigTech workspace, even if they have not been added to the workspace itself.

Supports Framework v9 only#

Platform v2 only supports the latest Framework v9, which comes with major improvements and new features over Framework v8. For a look into v9, read our Guide to SigTech Framework v9.

If you’re an existing user moving from Platform v1 to v2, please note that you won’t be able to use any v8 strategies in the new platform. If you’re using a version of the SigTech framework that’s older than v9, you may need to update your code to ensure that it runs and is compatible with Framework v9. The Migration guide has further instructions on how to do so.

GitHub integration setup#

To use BYOR’s enhanced GitHub integration, you’ll need a GitHub organization or enterprise account.

Your GitHub org members can be added as SigTech platform users using SigTech’s web-based account management features—just make sure users in your organization sign in to GitHub and SigTech with the same company email address.

You’ll need to install the SigTech GitHub App in your GitHub organization account. Please refer to our page on SigTech GitHub integration for instructions on how to set up the SigTech GitHub app and integrate your own repository with SigTech.

Note: To remove the SigTech GitHub App’s access to your GitHub repos, either uninstall the App from your GitHub org, or explicitly revoke the particular repo from the list of repos the App can access. Your code will remain in your GitHub org.

SigTech workspaces and GitHub repositories#

When you create a new SigTech workspace, the SigTech GitHub App creates a corresponding GitHub repository for the workspace, as well as creating a repo branch for each user added to your SigTech workspace. The app has permissions to read and write to any repo it creates.

When setting up a new SigTech workspace, GitHub Enterprise users have the option to set up branch protection for both the main branch and any user branches. The SigTech GitHub App configures any specified branch protection rules when it creates new branches of the workspace repo.

For further instructions, read our page on using the JupyterHub Git extension.

Research environment only allows one active repository at once#

Previously, v1 allowed multiple active repositories to be loaded in one research environment at once. However, cross-repository references would fail when deployed.

As a result of the improved integration with GitHub, Platform v2 only allows access to one active Git repository per research environment. This means you can only see the files in the workspace selected when creating the research environment. Furthermore, a full snapshot of your research environment, including all files from your workspace, is mirrored in deployment.

As a result, your executed code will always work exactly the same way as it did in the research environment. By design, having only one active repository at any one time should prevent deployment failures from occurring.

pip install no longer required#

Previously in v1, you had the option to pip install packages, which would run every time the research environment was started up (along with any other shell commands). In v2, pip install and other shell commands no longer run automatically.

This is because in the event of the user using pip install to upgrade a package used by the platform, strategy deployments would fail if the platform was using a specific version of the package.

In v2, you can still install pip packages using notebook cell commands. However, you will now have to run your notebook commands every time you start a new research environment in Platform v2.