SNAPSHOT SUMMARY FACT -

I have a situation where i need to take a monthly snapshot for payments , scheduled future payments and expanses and calculate variation at point of time.

We pay lump sump amount to external organisation on monthly basis as per agreement amount per FY. let say we agreed to pay 120 million per FY to particular organisation and it will be scheduled to pay 10 Million per month. we recieve claims (cost ) by organisation every months. how much it cost @ claim item level- then we aggregate that to a organisation level. what we need to see is how much we paid them and how much total cost we need to pay, to keep track of payments and expanses at point of time. we also want to see the variation and be able to compare at same time last year or year before. No payment is directly associated with the claim items. scheduled payments happens each month ( amount could change in future based on payment made against total cost) and we aggregated cost at organisation level, done in monthly basis to monitor if we over paid them if so we could stop or change amount at next schedule payment. at end of FY we reconcile and recover if over paid or make additional payment if it is under paid.
It is not just what has occured but also wants to track future schedule payments based on payment type, payment status. We need to calculate variation % interms of agreed payment to actual paid amount, variation on paid amount - total cost .I am more thinking of monthly snapshot (summary fact) with grain at organisation level.
New to Dimensional modelling just looking for a suggestions if i could take different approach or am i on a right track. If there are better way to handle situation like this .


thanks in advance for your suggestions.

Are you trying to develop an open analytic system or are you trying to write a report? That is basically what a periodic snapshot is. One could imagine having multiple snapshots over time, but what else could you do with it?

A fundamental principle of building an analytic foundation is to collect business data, such as these transactions, at the lowest level of detail possible in a proper star schema. This would give you greater flexibility to perform deeper analysis of the information that cannot be supported by an aggregate structure such as a snapshot.

After you have constructed atomic level facts (i.e. facts at the greatest level of detail), you can then consider aggregates built from the atomic facts to enhance performance or support specific business requirements.

Thank yo ngalemmo for your prompt response.
We are trying to write a report there isnt' any requirement for drill down at the momemnt. perodical snapshot, how much we paid? how much total cost ? how many payments are made to a date and how many remaining payment , outstanding balance and remaining payments within financial year and variance % etc. just monthly report to track payment/cost .