I really thought this was it.
I really had high hopes for this. The team seemed great, responsive and enthusiastic - the product was really good although was lacking in some ways, that seemed easily fixable. But they basic things never got fixed and now it's degrated further into something that is absolutely unusable. The bots can not answer a single question correctly without messing something up and most of the knowledge it doesn't access, even if you give it the most basic of knowledge.
Huge disappointment and has giving me plenty of headaches, both from a developer standpoint and a customer standpoint. Spent countless hours trying to get the bot to not be retarded, in some cases i was successful - but it was like playing whack-a-mole. One thing fixed another thing went bust, and the basic never got better.
Fatos_Answerly
Dec 10, 2024Hi Paradisian,
Thank you for sharing your detailed feedback. We deeply regret the frustration and challenges you've experienced. We understand how crucial reliability and accuracy are for both developers and customers, and we are striving to continuously improve our platform to meet those expectations.
It's important to note that our platform is operating as designed, but embeddings as a strategy for AI knowledge retention has presented challenges. As the volume of training data increases, the quality of responses can diminish, which is an inherent limitation of current AI embedding systems in general. We employ the most advanced embedding techniques available in the industry today, but the field has yet to see breakthroughs that fully address these limitations.
Your experience highlights areas where we need to improve, and we are committed to refining our tools and processes to ensure better outcomes.
We truly value your patience and willingness to work through these challenges, and we hope to deliver updates that align more closely with your expectations.