3 ms·
Ok focusing on context makes sense but I’d challenge the idea that LLMs inherently fail without it. Some teams have used fine-tuned models or hybrid workflows w
by emmtold 2y ago
Ok focusing on context makes sense but I’d challenge the idea that LLMs inherently fail without it. Some teams have used fine-tuned models or hybrid workflows with partial context to generate useful IaC snippets
- fasten 2y agoAgreed 100%. LLMs are doing solid job at generating IaC but in a context where the person who use them knows what he/she's doing. In our case, remediaiton means an extra level of trust, where your infra is already super sensitive.
- emmtold 2y agowe have used some tools to generate terraform code based on our unmanaged cloud resource for instance and it worked well..
- fasten 2y agoThe tools we are aware of will create a 1-to-1 mapping to some code, but very often with hardcoded values because they lack the full context of your infrastructure. This can lead to potential incidents in the future (broken dependencies / visibility). This is at least the way we are approaching it, and why we want to build this "deterministic" part first and then use it as context to the LLMs.