Your Digital Portfolio Assistant
My personalized chatbot isn't just a static FAQ engine; it's a dynamic guide to my professional experience and expertise. By leveraging both Botpress for the user interface and conversation flow handling and Stack AI for the text generation utilizing the OpenAi API, the chatbot offers rich, conversational interactions.
Two-Tiered Intelligence
The user prompt is first entered into the Botpress flow which then sends a request to the Stack AI API, which queries a vector database containing detailed information from my portfolio, blog posts, and more. The most relevant data is then selected and sent along with the user's original question and a customized system message to the Open AI API. The resulting reply is then forwarded back to the user.
What Makes It Special?
This dual-engine approach allows for future integration of more complicated flows in Botpress that can enable fewer calls to Stack AI and more nuanced replies. I have chosen to limit these flows in this bot to make it appear more conversational and to show off the power of Stack AI. It's like having a conversation with me, backed by all the data points that constitute my professional journey.
Limitations and Improvements
While these tools allow for streamlined development, they are not the answer for every chatbot. It's crucial to balance between efficiency, cost, and functionality to determine the best approach. In the future, I aim to explore optimizations, and possibly integrate more datasets and intelligence layers to provide even more precise and tailored responses.