Explore Your Dream Job With AI-powered Career Path Navigator : Data-Driven Career Guidance with Gemini & BigQuery
<Introduction | Overview>
The idea was to build a Career Counselling Support Portal that is a comprehensive web-based platform designed to help individuals discover and navigate potential career paths. The portal aims to bridge the gap between individual aspirations and the dynamic demands of the job market, offering a user-friendly interface and expert guidance to facilitate informed career decisions.
Here’s a brief idea about what we will currently build:
- Intuitive web-based UI
- AI Model fine-tuned by a prompt
- A cloud container to host the website
- What the reader is expected to accomplish by the end of this task.
We will also peek into the on-going journey of my project: the career-path-navigator. I aim to build a solution that leverages advanced data analytics and personalized assessments to provide users with tailored career recommendations, educational resources, and job market insights.
< Idea | Design>
Include high-level architecture of the solution. Outline the rationale behind the design and its impact on usability and functionality.
Here we will build a career-navigator app, that will have a frontend (index.html) , a backend(app.py), some requirements and docker containerization.
Apart from this,
- We will be analysing a csv based career-aspiration dataset and a employee job dataset, cleaning and converting it into jsonl format.
- Fine tuning a text-bison model to provide the predictions.
- Creating a RAG application that generates suggestions after analysing a users portfolio.
<Prerequisites>
Ensure that you have the necessary software and basic knowledge.
- Python
- HTML, Javascript
- GCloud
- Docker
<Step-by-step instructions>
Here we go through the building process:
Create a frontend that provides access to all the features currently available
In this case, it currently allows the user to enter a question to the ai model regarding any career advice s/he seeks.
Create a backend that makes all the necessary API calls and all the necessary functions to run the service
<Result / Demo>
<Other development>
.jsonl file created from csv dataset
model training in google collab
trying rag implementation with some relevant files