Tech Stack Forecaster

Tech Stack Forecaster: Stop Guessing. Start Predicting. An AI-powered platform analyzing 5 years of BigQuery data to forecast 2025 tech trends with 94% confidence.

Comment

Tech Stack Forecaster (AI-Powered)

Sheridan Datathon 2025 Submission > A data-driven platform predicting the future of programming languages using Google BigQuery & Machine Learning.

The Problem

In the fast-paced tech industry, students and developers struggle to decide what to learn next. Subjective YouTube videos aren't enough—we needed hard data to answer questions like:

"Is Java actually dying?"

"Is Rust just hype or a real career path?"

"What will be the most popular language in 2025?"

The Solution

Tech Stack Forecaster doesn't just show history; it predicts the future. We analyze millions of questions from the Stack Overflow Public Dataset on Google BigQuery and apply a Linear Regression Machine Learning Model to forecast trend lines for the coming year.

Key Features

Real-Time Data: Queries 5 years of live data from BigQuery (2019–2024).

AI Forecasting: Uses scikit-learn to calculate slope and projected growth for 2025.

Smart Verdicts: Automatically classifies languages as "Skyrocketing 🚀", "Plateauing ⚖️", or "Declining 📉".

Mobile-Ready PWA: Fully responsive design that looks native on mobile devices.

Tech Stack

Component

Technology

Frontend

Next.js 14, React, Tailwind CSS, Chart.js

Backend

Python, Flask, Pandas, NumPy

AI / ML

Scikit-Learn (Linear Regression)

Database

Google BigQuery (Public Datasets)

Cloud

Google Cloud Platform (GCP)

Setup & Installation Guide

Follow these steps to run the project locally.

Prerequisites

Node.js & npm installed

Python 3.9+ installed

A Google Cloud Service Account Key (service-account.json)

Backend Setup (Flask + ML)

The backend handles the BigQuery connection and the Machine Learning logic.

Navigate to backend folder

cd backend

Create and activate virtual environment

python3 -m venv venv source venv/bin/activate # Windows users: venv\Scripts\activate

Install dependencies (Flask, BigQuery, Scikit-Learn, Pandas)

pip install -r requirements.txt

⚠️ CRITICAL STEP ⚠️

Place your 'service-account.json' file inside the 'backend/keys/' folder.

If the folder doesn't exist, create it: mkdir keys

Run the server

python3 app.py

Server should be running at http://127.0.0.1:5000

Frontend Setup (Next.js)

The frontend visualizes the data with interactive charts.

Open a new terminal and navigate to frontend

cd frontend

Install packages

npm install

Run the development server

npm run dev

App should be live at http://localhost:3000

How the AI Model Works

Data Extraction: We extract monthly question counts for specific tags from BigQuery.

Vectorization: We convert time-series data into feature vectors.

Training: A Linear Regression model is trained on the last 5 years of data points.

Prediction: The model projects the trend line forward to predict the total count for 2025.

Classification: We calculate the coefficient (slope) of the regression line to determine the "Verdict" label.

Team

Built with ❤️ for the Sheridan Datathon 2025.

Anas Abu Mosameh - DevOps Engineer - Cloud Security at Sheridan College

Share this project:

Updates