Customer queries/datamining are currently developed in SQL. Each sample covers a particular error pattern, and they have 70+ queries ("criteria"). A lot of logic repeats itself in these criteria but is hardcoded in each criteria - the code is shared and not recycled.
The ambition is to reimplement the 70+ queries in Python instead of SQL. By using Python, it is expected that a number of winnings can be harvested:
- option to use version management
- The ability to (automatically) test our code
- better option to make reusable code (do not repeat yourself)
Python is also used in other solutions. Thus some co-production benefits are also gained by switching to using Python instead of SQL in search for errors.
The need is an experienced Python software developer that can help us migrate our datamining from SQL to Python.
In addition, the customer needs an architect who can help design the solution that the upcoming Python datamining will drive.
If the two needs - Python developer and architect - can be covered by the same consultant, then it's very nice. Alternatively, there must be two consultants.
- Solid experience with Design Patterns, Object-Relational Mappers, Virtual Environments and Object Oriented Development is a requirement.
- Healthy development habits are also a prerequisite: divide code into clear sections (use packages, modules, classes, functions appropriately), apply consistent version management, and experience with using tests.
- The consultant must participate in the specific work of migrating code from SQL to Python. It will take place in collaboration with people in the team, which will ensure two things: 1) the employees can help the consultant with business understanding of data; and 2) the consultant can help lift their Python level by code review, compliance with architectural principles, compliance with coding best practices, etc.). 2) means that the consultant must be prepared to - and good at - cooperating with and teaching other people.
- Can think all the way from database level to "application", ie. the individual queries written in Python. There must of course be some abstraction layers in the solution, such as Data Access Layer and Feature Layer.
August (early or middle)
Min. fem års professionel IT-erhvervserfaring.