Karamashth

What it does

Karamashth analyzes individual's profile, identifies the strengths, weaknesses, and career goals. By comparing individual's profile with other successful individuals in respective field, it generates a tailored roadmap for one's growth. Here Advanced AI models, including document AI, recommendation AI, and prediction AI, work together to provide actionable insights. There are two different paths to selects: "PDF doc gap analysis" and "Image gap analysis". The uploaded docs are treated as RAG thus gemini will have bounded context. For PDF analysis, I have used "gemini-1.5-pro-001" model with right blend of prompt engineering including few-shot examples, thus the response can come in HTML page with unique checkboxes for all the tasks of a given roadmap (in this case, "Technical manager to CEO"). Then for Image analysis, I have fusion the two models: "gemini-1.0-pro-vision-001" that analyze the images and find the gaps based on the prompt provided, however vision model is limited to vision only thus I again used "gemini-1.5-pro-001" model with old few-shot example prompt to get the HTML page with unique checkboxes task for Image analysis. And Karamashth offered to select different task and prepare unique roadmap (based on personal choices) then I called 3rd model "gemini-1.5-flash-001" to get the quick suggestion for given roadmap and task to accomplish. Last by not the least, Karamashth is tracking the progress graph of each roadmap to help one's aspiration!