openIDS Work Group - 9/26/24 - Minutes
Meeting Participants
James Bamberger
@Sean Bohan
@Jefferson Braswell
@Peter Antley (Deactivated)
Josh Hershman
@Ken Sayers
Mike Puchner
Michael Sullivan
Michael Schwabrow
Catherine Myers
@Greg Williams
@Nathan Southern
Andy Mielke
Michael Payne
Andrew Regis
Jenny Tornquist
Cory Isaacson
Rob Clark
Opening Comments
Group lead Jenny Tornquist (JT) reviewed the agenda for this meeting:
Confirm the new working group name – Open Insurance Data Standards (OIDS) at openIDL
Decide on the Base Model based on the two models presented (pros vs cons): Cloverleaf Analytics OMG model Lloyds CDR model
Based on the decision in #2, we should identify any limitations on the following: Legal reliance on ACORD (copyrights)
What areas will or will not be OPEN SOURCE? When do we want to introduce carriers to the Working Group?
Discuss LF role in Governance and change management
Group Name Review
This was discussed - the group has been renamed Open Insurance Data Standards at openIDL (OIDS) Work Group. Simultaneously Werner Kruck at AAIS has been working on a broader rebranding for openIDL that will require following a host of other procedural steps; once this occurs, ‘at openIDL’ will be dropped. Josh Hershman will keep everyone posted on this process and where we stand.
Decision Making Process - Insurance Model
Two reviews took place of base data model options for this work group - one of the Cloverleaf Data Model (presented on Mon. 9/9) and one of the Lloyds Data Model (LMG CDR, presented on Thu. 9/12) - and Ms. Tornquist opened this topic up to the group for review. The following discussion took place:
Mr. Sayers affirmed the need to find a starting point for the group to get its work underway and the fact that the group needs to make a final choice.
Mr. Isaacson feels the Lloyds CDR is not ready and should be ruled out. He also feels that, by extension, the Cloverleaf is a great starting point - the Cloverleaf team is willing to open source it and work with the group as needed, add to it as needed, etc. The Lloyds CDR can potentially be used in the background as a reference point.
Mr. Antley is also in favor of starting with the Cloverleaf. Mr. Sayers agreed that adopting this makes sense, and stressed the need to define what open source means in this context, and what is the dependency on ACORD and is there any risk present with it?
A soft vote took place to start adopting the Cloverleaf model in group - no one in meeting expressed any opposition to using the Cloverleaf model; Mr. Isaacson pointed out that it is very close/similar to the reThought data model. It is based on OMG 2014 P&C Insurance Models.
Open Sourcing Discussion
Mr. Isaacson then raised the question of what license needs to be used for this work (GPL or Apache). Per Mr. Bohan, Apache 2.0 affords the greatest flexibility from an IP perspective. Mr. Isaacson agreed and there were no objections from the group. The Cloverleaf members present also concurred.
In terms of open sourcing: it was stressed that the Cloverleaf team will need to donate the analytics model to the foundation under the Apache 2.0 license as Intellectual Property.
Mr. Bohan offered to set up a meeting between the Cloverleaf representatives and Linux Foundation attorneys to discuss the logistics and implications of this.
Mr. Sayers stressed that two separate Cloverleaf models are pertinent, here:
The simplified transmission model
An internal model against which querying is done
Cloverleaf response: this whole effort benefits their business operation; they have also internally discussed open sourcing the model and are in favor of doing so. A standardized way to communicate between insurers is a significant asset for them.
Mr. Isaacson re-stressed that we’re looking at a donation of the two above models.
-It was requested that an Atlassian (wiki) page be created for this working group - where contributors to the WG can go and edit it and leave comments. Sean and Nathan agreed to set this up.
Legal Questions About ACORD
Ms. Tornquist noted that Cloverleaf has relied ACORD to some degree in their work she raised the question of whether there are any legal conflicts/issues here.
Per Rob Clark: Cloverleaf previously (about 7 years ago) used the ACORD data model to some degree, but abandoned it, as it proved too complicated for what they were doing - too granular, etc.
The only elements Cloverleaf retained from ACORD were the attributes they collected - year, make, model (on auto), etc. Aside from this they have retained nothing from ACORD; Mr. Clark has already consulted with the Cloverleaf attorneys. Cloverleaf is not using any material that is proprietary to ACORD.
AAIS acknowledged this but asked Cloverleaf for formal legal documentation to this end. Cloverleaf agreed to produce a letter that AAIS can share with its attorneys.
Mr. Isaacson suggested that the attorneys be asked to present a letter proving that Cloverleaf independently developed their model without relying on other sources - this may be easier to obtain than a legal document stating no connections to ACORD’s data model. Cloverleaf agreed.
Introducing Carriers to the Working Group
AAIS has two carriers that want to participate in this group and asked when it makes sense to bring them into this working group.
All agreed that it makes sense to get everything buttoned up (legally and procedurally) before carriers are brought in. A working prototype should ideally be finished and ready - with software that is ready to roll out. A proof of concept and multiple use cases should be ready before carriers are brought in. This means having critical mass in the components.
Action Items
Sean Bohan - set up a meeting between the Cloverleaf representatives and Linux Foundation attorneys to discuss the logistics and implications of open sourcing their data model.
Cloverleaf Team - Consult with your internal legal representation to request a letter stating that you independently developed your model without relying on other sources; share this, in turn, with AAIS.
Sean Bohan and Nathan Southern - Set up editable Atlassian wiki page for OIDS working group.