The big difference with smaller models, is that you can more easily inject your knowledge into them, confirms Christian Stocker #LIIP in sharing the story of open source RAGs from crawler to LLM to production app in one of the first useful public apps with #Apertus

Try it at https://oss.zuericitygpt.ch

Tip: be specific in your prompts, or be prepared to wait. (Kind of like at an actual info-desk..)