Project and track management

Project Management on the Allonia platform works with two levels called Project and Tracks. As the Project is the top layer workspace, the Track let you manage more specialized sub layer of your project (as for a specific feature or technical environment for example).

Several tracks can be included in one project. Depending on your needs, you will be able to assign different roles to users on different tracks.

Projects

project view

Create a new project

  1. Click Project Manager → Create new project.

  2. The project creator will automatically and by default have the role of the project manager.

project create

When creating a new project, a first track by default is automatically created and the project creator is automatically redirected to the configuration form of this first track.

The project creator must configure the first track created by default by specifying the following informations:

  • Track name

  • Compute configuration

  • Notebook compute environment type

When creating the project, the creator has the possibility to add users to the track by default and assign them one of the following roles: Track leader, data scientist or data engineer

Manage project manager

You can change the project at any moment from the "Project management" view

project manager

Archive a project

If your team has completed a project, you can archive it so that it does not remain in place if it is no longer needed. An archived project will no longer appear in Allonia, but you can continue to view archived issues in read-only mode, through direct links or mentions in other projects.

Tracks

Allonia track is a sub-entity of the project used to organize the project in several steps and customize each step with adequate resources in users/roles and resources.

track view

Create a track

  1. Click Project name → Create a new track.

  2. Choose the Track name

  3. Choose Compute configuration

  4. Choose Notebook compute environment type

track create

When creating the track, the creator has the possibility to add users to the track and assign them one of the following roles: Track leader, data scientist or data engineer.

Manage track users

User can only be managed at track level, before that, any user have to be added by the organization administrator inside the organization.

  1. Project management → Click the target project→ List of tracks.

  2. Click “Actions” of the target track

  3. Click Manage Users

  4. Add the new user from the user list and assign him a role

track manager

Roles

Roles responsibilities

Role Description

Tenant Admin

Is added by the Client during account creation Manages the client account: Billing, Data governance, Licences and Users

Project Manager

Is added by Admin during project creation (by default himself) Manages the Project: Tracks, Role-Users, Resources, Costs

Track Leader

Is added by Project Manager during Track creation (by default himself) Manages the Track: Data access rights, users, publications…

Data scientist / Data Engineer

Is added by Project manager or the track leader Contribute to Track objectives by performing tasks: improve a Dataset, Model or Pipeline

ML Ops

Is added by Project manager or the track leader Contribute to Track objectives by performing tasks: deploy and manage jobs, user-services a Dataset

Roles actions

Action Tenant Admin Project manager Track leader Data scientist / Data engineer ML Ops

Create project

Yes

No

No

No

No

Add new track

Yes

Yes

No

No

No

Add user

Yes

No

No

No

No

Edit user

Yes

No

No

No

No

Manager user roles in project

Yes

Yes

No

No

No

Manager user roles in a track

Yes

Yes

Yes

No

No

Roles ressources

Role Object scope

Tenant Admin

All objects in any projects

Project Manager

All objects in the projet (potentially multiple tracks) he manages

Track Leader

All objects in the track he leads

Data scientist / Data Engineer

All objects in the track(s)

ML Ops

All objects in the track(s)

Roles access

Action Tenant Admin Project manager Track leader Data scientist / Data engineer ML Ops

Project view

Yes

Yes

Yes

Yes

Yes

Factory

Yes

Yes

Yes

Yes

Yes

Launchpad

Yes

Yes

Yes

Yes

Yes

Data catalog

Yes

No

No

No

No

Administration

Yes

No

No

No

No

Documentation

Yes

Yes

Yes

Yes

Yes