Questions can all be answered with claims tables developed in this group
Personal Auto, Commercial Auto, Personal Property, Commercial Property, Business Interruption, Flood, Other
Have stat plans for all
PA to look into how commercial property is loaded
what stat plan would it come in on for Flood?
Fields for commercial vs personal auto
(DH shares screen for commercial vs personal auto)
Day 1, 2, 3 type fields - only thing in stat files today is day 1, gone thru and ID'd personal, commercial, both and separate
PA, JM and DH working on dividing up fields by field type
DAY 1
for Day 1, apply to both personal and commercial (DH goes through the spreadsheet)
MN - keep in mind, right now focusing, can we look at some fields holistically, as we expand we want to make sure same meaning is across the board, not need to go back and make adjustments
for data model for auto, can apply to mult lines
can only have one stat agent in a given state per company
DH - only diff they see is Mass and Texas
DH - as we think about putting together data structure - how many diff tables will be needed - where is the overlap between personal and commercial auto
may have pricing terms within personal but not commercial, may use alt companies on either side too (differentials in price)
Driver attributes are personal only
most coverage rows apply to both
NJ threshold tort limit is personal only, combined first party benefits is personal
Catch up on AWG documentation
Peter to find raw copies of stat plans and start implementing changes
dont want to onboard a carrier and not have documentation for what and where everything is
cobble together our own stat plan
one stat plan for personal and commercial auto? can we build a strong foundation for the data we are working with? Suggest we have a stat plan for auto, for commercial auto and full documentation for standalone things - as time goes on we can add more/not diverge
Begins to intro notion of wherre do you put things into more than one table - ease into the topic
Personal and commercial auto are TWO different insurances - logical to separate (esp at the DB level) - should be true with Day 2, Day 3 fields as well
do you have something consistent with auto, things that are variations - data modeling discussion
keep it straight for people using it - make it easy to onboard people
How big of data should we make the openIDL pipelines support - how much in a month or how much to store?
When you do a batch to load, an average 1/12th of a year or so, depends on how many customers and what line of business, all lines for a given month?
when we have a job that needs to be within a threshold like end of month submission, how big/how much?
is 10MM a month per carrier a decent sample size we are sure we could handle - DH says to be safe double it
20-30MM, robust loading, anyone who can handle 30MM should be good
as soon as we get more test data set up, more lines, want to try to get some regulators and see some sample data calls
data catastrophe call - get EPs in Feb or Mar, putting data in, doing catastrophe in MISS, could do these calls on any using data standards, look at historic data calls this spring
want to reach out and speak with Regulators, produce value for them