Inquisiv.

Streamlining the Drop shipping process of market research and branding for e-commerce stores

Comment

Inspiration

Dropshipping entrepreneurs face an oversaturated market and struggle to keep up with rapidly evolving trends. Conducting market research and creating ads takes weeks, which delays their ability to scale. We wanted to streamline this process using cutting-edge technologies, enabling dropshippers to save time and make data-driven decisions.

What it does

Inquisiv automates the dropshipping process by:

  1. Product Discovery: Uses web scraping and machine learning to identify top-trending products based on reviews, ratings, and sales data.
  2. Sentiment Analysis: Employs NLP techniques to analyze product reviews and predict customer satisfaction.
  3. Ad Generation: Integrates APIs to generate customized ad content for selected products, reducing time spent on marketing efforts.
  4. Insights Dashboard: Displays actionable insights, such as product trends, customer satisfaction scores, and data visualizations.

How we built it

  1. Languages: Python, JavaScript, HTML, CSS, Jupyter Notebook.
  2. Backend: Flask and REST APIs for data handling and processing.
  3. Frontend: Built with React, React Bootstrap, and TanStack Query for a seamless user experience.
  4. Web Scraping: Used BeautifulSoup and Playwright to gather product data from e-commerce platforms.
  5. Machine Learning: Leveraged libraries like Torch, sklearn, and NLTK for sentiment analysis and trend predictions.
  6. Visualization: Employed Matplotlib, Seaborn, and SHAP for data visualizations and insights.
  7. Development Tools: VS Code, PyCharm, Google Colab, Figma, and the Northeastern Research Cluster for high-performance computing.

Challenges we ran into

  1. Data Acquisition: Scraping large amounts of data while adhering to platform policies.
  2. Model Training: Fine-tuning the machine learning models for sentiment analysis and accurate predictions.
  3. Integration: Combining web scraping, backend APIs, machine learning, and frontend into one cohesive platform.
  4. Time Constraints: Delivering a functional prototype within the hackathon timeframe.

Accomplishments that we're proud of

  1. Successfully implemented NLP-based sentiment analysis for product reviews.
  2. Created a fully functional prototype with real-time product recommendations and automated ad generation.
  3. Built a visually appealing and intuitive interface using React.
  4. Leveraged advanced ML techniques like gradient descent to improve prediction accuracy.

What we learned

  1. The importance of efficient collaboration and time management in a high-pressure environment.
  2. How to combine multiple libraries and technologies to create a unified solution.
  3. Advanced web scraping techniques using Playwright and BeautifulSoup.
  4. Leveraging tools like SHAP for interpretability in machine learning models.

What's next for Inquisiv

  1. Secure funding: Use funding to gather and buy higher end cleaner data in larger sets
  2. Launch the Product: Gain at least 100+ active users by 2025.
  3. Expand Features: Add more APIs, such as Canva and Adobe, for advanced ad customization.
  4. Optimize ML Models: Further refine sentiment analysis and product recommendation algorithms.
  5. Global Expansion: Secure funding to scale internationally by 2027 and partner with companies like Kaoldata and WinningHunter.

Built With

Share this project:

Updates