This is a small research project that can determine which resources work best for job searches and help me find new job
-
The competition among developers is quite big and the candidate must be something different, as I thought such a project could attract attention
-
I was curious to try embedding api from OpenAI
-
I will be able to see and share with other applicants which job search resources work best
-
I do not need to send my cv's
-
I create redis index and populate it with 2 embeddings "backend developer" and "product manager" (if you didn't get any resume just enter one of those)
-
Each time user enters desired position on the website I send request to OpenAI, calculate embedding for it and calculate cosine similarity between each embedding in index and one that was entered
-
I get most similar embedding from known embeddings and if similarity is more than 85% users get's my cv, otherwise not
-
I store the fact of querying API and utm_params (if any)
-
I keep the text of user entered position and the similarity score (no utm's or ip)
-
Downloads for each cv with utm's (so I can evaluate conversions)
-
For the website analytics I use simple analytics
Since I am looking for a job as a product manager, I couldn't do without the analytics dashboard
Analytics consists of two tabs
- Web analytics (taken from simple analytics API)
- App analytics (taken from data collected from my API)