The Summer of Data
If has a tradition of hiring a fair number of interns during the summers. This is an opportunity for us to develop relations with potential future employees and for the students to get a glimpse of work life and test their skills in practice.
In this summer’s cohort we had three talented young women from the Norwegian University of Science and Technology (NTNU). Isabel Thy-Van Lam, Jenni Sveen Olsen and Ingrid Renolen Borkenhagen are starting their 3rd (Isabel) and 4th (Jenni and Ingrid) year of studies in NTNU, specialising in industrial economics and technology management and engineering & ICT.
Approaching the summer, their knowledge about the insurance industry was limited to owning a travel insurance combined with some assumptions of a hierarchical, male dominated, formal organisation (the reality turned out to be quite the opposite!).
Although all three had solid technical understanding and some programming experience (mainly in Python and Java) going into the internship, none of them had real-life experience of working with large data volumes. This was a chance to learn new skills in an area with growing demand for competent people.
The work tasks during the summer revolved around the theme of development of If’s data products — improving data quality and usability of If’s common data assets for analytical use cases. For those interested, there is a separate article about our common data products here.
As the data platform, language (SQL) and If’s data assets were largely unknown to Isabel, Jenni and Ingrid, they started the internships by doing courses in the Analytics Academy — If’s internal offering for data & analytics related learning.
“The introduction courses offered by the If Analytics Academy as well as coaching, helped us get going.” — Isabel & Ingrid
After a short and an intensive training period, the interns were thrown to the deep end and started to work on actual development tasks. One specific task was to research the quality of data in If’s Claims Data Product. More specifically, could they find and document discrepancies between data in our enterprise data warehouse and the source systems?
The quality assessment work began with simple queries to get familiar with the data and to discover potential challenges but quickly evolved to more complex solutions that were tested and optimised. Towards the end of the summer, the different queries had developed to a dynamic query that is both efficient and easily replicable to be used on other data assets. The solution is described below.
The function of the code below is to take the input parameter of a lookup table and perform a series of queries to detect missing/faulty column values of interest. Afterwards these inconsistencies populate a target table that can easily be accessed by a user. The end results can also be accessed through a BI solution built on top of the target table.
-- The purpose of this code segment is to iterate through all the query strings and execute them one by one-- The results of the queries will then be inserted into a volatile result tableBEGIN DECLARE QueryString VARCHAR(1000); DECLARE counter INTEGER; DECLARE rowCursor CURSOR FOR SELECT Select_Query FROM VT_Select_Query;--The rowCursor will select each row in the VT_Select_Query table OPEN rowCursor; SET counter = 0; WHILE (counter < (SELECT * FROM (SELECT Count(*) AS counter FROM VT_Select_Query) AS dt)) DO
--Inserts the selected row into a parameter FETCH rowCursor INTO QueryString; --Executes the string and inserts the result CALL DBC.SysExecSQL('INSERT INTO VT_QualityTable' || QueryString); SET counter = counter + 1; END WHILE; CLOSE rowCursor;END;
During the summer several areas of improvement potential were uncovered that are now being worked on. These improvements will have a direct impact on the accuracy of If’s data driven solutions built using the data.
“It is motivating to work on real-life problems and create solutions that end up in the production environment.” — Jenni
If as a company turned out to be a pleasant surprise as well. The team where the interns worked consisted of 50/50 women and men and a wide diversity of cultural backgrounds. The culture was informal, helpful and inclusive.
Much of the code that Isabel, Jenni and Ingrid developed will continue to live in If and create value while they resume their studies. We in If have been fortunate to be able to work with you. Thanks for working with tech that matters.
Good luck with the rest of the studies. We hope to see you again!
If hires summer interns for technical roles in all the Nordic countries. If you are interested, please get in touch with us at www.if.no/intouch.
Thanks to Emma Linnea Wiström for help with the article and for the excellent support and coaching of the summer interns.
