5 tips for dealing with IT

Getting data science done with your IT team

Analysis
R
Author

Dean Marchiori

Published

November 18, 2024

No plan survives first contact with IT

In most organisations the single biggest challenge you will face as a data scientist is dealing with IT (Information Technology / Technical Support) to support and enable your data science workflows in Production.

Why?

It’s not really anyone’s fault, it’s just a lack of alignment.

IT teams don’t care heaps about enabling 1% of users to run wild with Open Source technology or free software or weird programming languages.

They care about managing risk. They want to enable 99% of users to do their job safely and productively without interruption. They want you to succeed, but they deliberately want friction in the change management process to reduce the risk of breaking changes.

Tips to getting stuff done with IT

  1. Have empathy: Understand what their role is. Learn how little you matter to them. Figure out what would make them look good. Find your common enemies and work together. Don’t get frustrated at them. It’s a feature not a bug. You just want different things. You have no idea the shit they have to deal with every day.

  2. Political alignment: You can go into battle with some poor IT manager with a million bigger fish to fry, or you can try to influence your own manager. They are actually motivated to ensure your success and will be able to influence IT in a more effective way on your behalf.

  3. Find supporters: Okay sometimes you just need a favour. Find out who likes you and if they are willing to give you a hand now and then. Reward generously with coffee and thanks. Swing by their desk, ask about cricket or what flavour of Linux they are running at home.

  4. Find (safe) workarounds: There are areas of grey where policy meets reality. Live in those areas. IT staff are smart technologists, they exploit this grey area too. Just do it ethically. If you org has actively blocked something and have no risk appetite to do something, then don’t bypass it. Otherwise, sometimes it’s better to beg for forgiveness than ask for permission.

  5. Buy commercially licensed software: If you go to your IT manager with arguments like: R/python is free, its open source, it wont cost anything, I can manage it myself, everyone cool on the Internet is using it, SAS/STATA are crap - then you wont get far. IT don’t care about free and open source. They want to know the software is supported so they can get help when inevitably users call them for help. They want to know the legal terms and conditions to keep legal happy. They want to ensure security vulnerabilities are patched (not by you). They want to know how it will integrate with the enterprise identity management solution. How will users be managed and authorised. They want to manage software in-line with the other billion apps they support and not by some exception because of some nerdy business user (you). Consider buying commercially licensed software like Posit Connect or Workbench. All of these problems go away and its just another piece of software. The argument then just devolves into who pays. And guess what? Everyone in a big organisation has it in their head that software licences are expensive so the sticker price won’t really surprise anyone.

Bonus tip: Share success: Don’t be a jerk and twist arms in other departments and get what you want then take all the credit yourself. Share success. Present to your IT team’s monthly meetings, include them in yours. Develop a ‘partnership’. They might have to be the ‘gatekeepers’ but they work for the same company because they want to make an impact too. So recognise it and appreciate it. You can’t do impactful work without them.

Want to read more? Sign up to my mailing list here for weekly emails and no spam.
Looking for a data science consultant? Feel free to get in touch at wavedatalabs.com.au