We have just launched a new project called Sextant Fleet. It is part of a Dutch government backed community project called dawo and is aimed to be an MDM solution to roll out secure laptops and manage them from an easy to use interface much like Intune.
Some of the mean features are:
near zero touch enrollment of laptops using PXE boot and MOK keys.
Policy and setting management
Pull based updating using Comin
Security compliance with Dutch government standards (which are mostly derived from ISO27001)
AI disclosure: We encourage the use of LLM’s as an assistant. In our policies we clearly define that using LLM’s in this project is fine but that a human is always reponsible for the shipped code. AI is used for refactoring, testing, building pipelines and writing MVP’s.
This looks amazing! It even has SSO! It could be really useful for small companies.
Is this going to be adopted by municipalities? Who’s pushing this? In my experience companies end up with Microsoft because they have an amazing sales team, spite of everyone being against it and having better solutions out there, upper management gets swing to the MS crap.
Seeing a government pushing for this is quite incredible. I wish you the best!
Cool stuff, as a person thats living in the netherlands, this is particularly interesting. Though i have to once again ask the question, have any LLMs been used during the development of this project? NixOS discourse has a policy which requires disclosure of any significant LLM use.
Thanks for the heads up! I’m in discussion in PM’s with Hexa on how to properly disclose the usage of LLM’s. As I’m quite new here (created my account yesterday and there was no linkage to this policy on the welcome to discourse page so I missed it…)
We greatly encourage LLM use but mainly as an assistant for stuff like writing documentation, refactoring, building POC’s and reviewing/testing code. As in our opinion it is yet another tool in our toolbox that we should not shy of using where it is appropriate. All of the code needs to be signed of and taken responsibility for by humans.
Thank you very much! Currently NixOS is being piloted at a few municipalities. Sextant is part of the second phase and will be used from September onwards. Other parts of the government are showing interest as well.
Hopefully it will stick. Especially for managing large fleets the power of NixOS is unmatched. The IT teams seem to recognize this as well which is hopefull.
We greatly encourage LLM use but mainly as an assistant for stuff like writing documentation, refactoring, building POC’s and reviewing/testing code. As in our opinion it is yet another tool in our toolbox that we should not shy of using where it is appropriate. All of the code needs to be signed of and taken responsibility for by humans.
Thank you, personally this is going to be the last interaction I have with this project sadly. My issues with LLMs have almost nothing to do with the tools practicality or responsiblity. I take issue with the mass scale plagiarism that occured during training, the societal, environmental and ecological impacts. “Just another tool” is ignoring all of the above.
But nonetheless, i wish you luck with your project.
Thanks for your reply. If you ever change your mind (not about AI but about the project). Please feel free to give it a spin. As the project is still quite young any input is valuable. Nothing is worth more than a good set of human eyes.
