I feel like the ai code generation companies such as cursor and Windsurf have completely ignored the world of embedded software development. Is there anybody in this ecosystem who has been able to successfully utilize AI tools to to develop embedded software.
If yes I would like to see specific examples of how it has been useful as well as what tools were they using please. TIA.
PS: Feel free to mention any AI tools that are helping in hardware development overall
You can try Github Copilot. It can handle tasks like writing boilerplate firmware, peripheral setup code, simple ISR templates, FreeRTOS tasks and HAL-based initialization.
Flux.ai is almost good for PCB designing
Like they said, github co-pilot is pretty useful
I’ve have good luck with Claude Code for small programs…but I generally have to have other AIs troubleshoot/help too (perplexity, chatgpt, sometimes Poe if I’m wanting options). I will say that the outputs I’ve gotten lately are about 10x better than 1.5 years ago, so that bodes well 
how do you guys use co-pilot when it does not have a hardware context. Maybe still workable for application layer, but any bare metal driver/BSP and related software, how do you do it?
I haven’t (yet!)
If you give it boilerplate code it should generate useful suggestions…and if you supply the missing context that you mentioned it’ll work great (though for implementations that have hardware-specific IDEs those will probably be better, especially if you are used to them…maybe cursor.ai would be helpful there as well)
I kicked your question to chatgpt and it has some good notes about the limitations…this is how I use it effectively. If I’m inexperienced in someth9ing I’ll kick that questions to one, and use others to verify/proof the suggestions. Claude thinks mostly the same