An smart internet surf engine powered by open source ai models.
→ Step 1: Receive the User's Question
The process starts by capturing the user's input question.
→ Step 2: Retrieve Top 6 Search Results
The question is passed to our Search API, which returns the 6 most relevant links.
→ Step 3: Scrape and Store Context
Text is scraped from each of the 6 links to compile necessary context information.
→ Step 4: Generate a Detailed Answer with Llama 3.1 70B
The user's question and the gathered context are sent to Llama 3.1 70B, which streams back a comprehensive answer in real time.
→ Step 5: Suggest 3 Follow-Up Questions with Llama 3.1 8B
Finally, Llama 3.1 8B is used to generate three additional questions that the user might ask to explore the topic further.