Totally get the reliability issues you've seen - getting LLMs to output the right response 80% of the time on these kinds of tasks is not that hard, the last 20

Totally get the reliability issues you've seen - getting LLMs to output the right response 80% of the time on these kinds of tasks is not that hard, the last 20% is the real challenge. On trying us out - ping me an email at henry@superflows.ai