The vector DBs and companies like OpenAI will eat up such small solutions very fast. They all have a great advantage already in the RAG space. So, better to think about creating something else other than building yet another RAG solution. But I feel like there is still a lot of gap in evaluation part, so watch out for that space.
AI is not a product, nor is RAG, they are at best algorithms. There is no moat at this low level, the field is rapidly evolving.
To differentiate, build a product for an industry that includes AI/RAG features. This requires expertise and experience in said industry to build something useful or competitive.
The vector DBs and companies like OpenAI will eat up such small solutions very fast. They all have a great advantage already in the RAG space. So, better to think about creating something else other than building yet another RAG solution. But I feel like there is still a lot of gap in evaluation part, so watch out for that space.
AI is not a product, nor is RAG, they are at best algorithms. There is no moat at this low level, the field is rapidly evolving.
To differentiate, build a product for an industry that includes AI/RAG features. This requires expertise and experience in said industry to build something useful or competitive.