v0.13

This new release allows you to fully leverage the Pivot added recently, by introducing the core bricks to build workflows

• New workflow tab added
• Create scenario by scenario workflows, based on decision outcomes
• Automatically route decisions to a specific case manager inbox
• Add to an existing case with the same pivot or systematically create a new case.

v0.12 - Pivot

v0.12 is a major release, providing the foundations of a key new feature : pivot.

You can now define a pivot in your datamodel. The pivot will allow you to group decisions (and in the near future, other features) by identical pivot to get global view on related alerts. For ex, you can set the user ID as a pivot => all future decisions can then be grouped by user ID, providing a bird eye view on the risk associated with him/her.

Backend :

• Feat : Creation of the pivot and storage of the pivot for each decision

Frontend :

• Feat : Create a pivot in the data model page
• Feat : Group decisions by pivot
• Feat : filter rules within a scenario
• Feat : Explore data model with a new view (available in "your data" tab)
• Fix : dropdown when selecting the field for a date

V0.11 bugfix

This release contains minor bugfixes and dependency updates.

v0.10 Fuzzy matching & wildcard decision API

**Frontend : **

• Feat : string fuzzy matching available from the builder. Two modes are available, check the documentation here
• Feat : Explicit error message on CSV/ batch ingestion timeout. Timeout is currently set at 50 sec.

Backend

• Feat : Decision API now allows to list all relevant scenarios for this object type, and create decisions for them. Scenario ID is not required in this case. Check the documentation here
• Feat : Decision API allows to query and filter all past decision. Multiple filters available
• Feat : Query scheduled decisions. Provides an alternative to set up an S3 to get scheduled decision

v0.09 What you see is what you get :)

We've added a really useful "view your calculations" when checking a decision within Marble's front.

Frontend :

• Feat : you can now view the detail of the rules for each decision. Two modes available : raw (shows the rules as designed in the scenario) and contextual (shows the real values used for eahc calculation).
  It makes things a lot easier to debug a scenario or check how a decision was calculated
• Refacto of outdated components (decision viewer, variable box...) to speed up future additional feature developpment
• Feat : display api key prefix in settings tab
• Fix : list value deletion bug

Backend:

• Feat : added fuzzy match operators back (front next week)
• Feat : Sentry monitoring for cron jobs
• Feat : auto retry ingestion in case of conflict error
• Feat : add audit trail on list value editing
• Fix : CORS settings work with open source version (for batch upload, decision download)

v0.08

Backend

• Feat : store rule calculation on each decision
• Fix: index on ingestion leading to poor performance
• Fix: CORS
• Drop clear API keys in DB, only use hashed values

v0.07 bugfix and small improvements

Small release with fixes, working on something big behind the curtains :)

Frontend:

• Fix: comment_added sur el case manager
• Fix: styling
• Fix: cap name on tables

Backend

• Feat: count(field) now counts all rows matching the filter, and not only the ones where the target field is not NULL
• Fix: timeout extended to 60 sec when sending files
• Fix: Error 500 when login as a Marble admin
• Security: fix allowed origin CORS

v0.05

Marble is available as a docker !

This release contains :
Backend

• a fix for CSV ingestion

Frontend

• the ability to open decisions in a second browser tab, without losing the filters

The major part of the release is the availability of Marble as a docker that you can self host on your computer or your own cloud.
Check the documentation for details

v0.04

Major version bump with unicity management in DB to ensure adequate performance of agregates.
First release with a decision to case automation

Frontend :

• fix: error in decision tab of a scenario
• feat: improve user auth error handling
• feat: improve data formatting (date, number, links…) in decision viewer and case manager. You can now add an clackable URL to each object, for use in the decision viewer and case manager (helps sending the operators to the right page on your back office for eg).
• feat: better instruction on how to deploy a scenario following the v0.03 performance refacto
• feat: add role in user info (top left of the interface)
• feat: new logic when opening an existing scenario (live if any, or most recent iteration)
• data model tab: tables appear collapsed by default

Backend :

• feat: automate decision to case creation (new field decision_to_case_inbox and decision_to_case_outcomes on the scenario (models, dto, db).
• feat: unicity management : to ensure performance and consistency, you can now define "unique" fields within each table. Those fields can be used to create link between tables and avoid duplicates.
• feat: store hashed value of generated api keys
• fix: bug fixes on DB field read
• reduced sentry sampling rate (backend)

If you are self-hosting Marble, you may have to run a migration script to ensure unicity for object_id and all other unique fields before this version bump.

V0.03 bugfix

Bugfix release

• 1st login attempt fails fix
• Better observability on Sentry (front)
• Agregate frontend overflow correction
• Bugfix on decision access that was partially broken in 0.03