diff --git a/src/04_Appendix/tooling-evaluations/langchain-haystack.tex b/src/04_Appendix/tooling-evaluations/langchain-haystack.tex index dd34c7294207fd9b54f76d4b860d14520973886b..9a01c6c59ffc4a31ab2be35e3a1840e9eec22497 100644 --- a/src/04_Appendix/tooling-evaluations/langchain-haystack.tex +++ b/src/04_Appendix/tooling-evaluations/langchain-haystack.tex @@ -21,7 +21,7 @@ It also has built-in support for a variety of NLP tasks. - The key feature \href{https://docs.haystack.deepset.ai/docs/pipelines}{pipelines} is the counterpart to the LangChain chains. + The key feature \href{https://docs.haystack.deepset.ai/docs/pipelines}{Pipelines} is the counterpart to the LangChain chains. Another key feature are the \href{https://docs.haystack.deepset.ai/v1.25/docs/agent}{Agents}. Particularly interesting is the \href{https://docs.haystack.deepset.ai/v1.25/docs/agent#conversational-agent}{Conversational Agent}, which simplifies the usage of the LLM in a chatbot scenario, similar to the LangChain ConversationChain. The Conversational Agent also handles the memory, so the LLM can answer context-based requests. @@ -30,7 +30,7 @@ As it is less flexible than LangChain, Haystack is easier to get started with. \subsection{Conclusion \& Decision} - Of course, in this evaluation, only a small subset of the available concepts were discussed. + Of course, in this evaluation, only a small subset of the available concepts and tools were discussed. However, most of the concepts can be found in both frameworks. That's why I focused on evaluating the concepts, that are important to this project.