Skip to content

AI-Hobbyist is a platform that helps users generate problem statements, clean datasets, and train AI models using a no-code approach.

License

Notifications You must be signed in to change notification settings

Neelkanth-khithani/AI-Hobbyist

Repository files navigation

AI-Hobbyist (Part A)

Many people see problems in the world but struggle to articulate them in a clear way that can help create solutions. They don’t know how to turn their thoughts into useful ideas. Using AI to solve these problems is also hard because of messy data and the need for technical skills. AI-Hobbyist helps users articulate their ideas by using stories, news, and PDFs to create problem statements. It also cleans messy data by fixing errors and missing values. With a simple, no-code approach, users can train AI models, check results, and download them, making AI easy to use for real-world impact.

Workflow Diagram

Workflow Diagram

The workflow consists of the following steps:

  • Input Sources: Users provide input in the form of problem stories, news articles, or PDFs.
  • Preprocessing & Feature Extraction: The system utilizes NLTK and Simple RAG to analyze text and generate a word cloud for key insights.
  • Problem Statement Generation: Natural Language Generation techniques and Meta LLaMA 3 refine and generate structured problem statements for dataset creation.
  • Dataset Preparation: The uncleaned dataset undergoes automated cleaning to ensure quality and consistency.
  • Model Training & Optimization: Users train machine learning models using a low-code/no-code approach, and results are evaluated for further optimization.

Use Cases

AI-Hobbyist can be applied in various domains, including:

  • Sustainability & Climate Research: Identifying pressing environmental issues from news and reports to generate structured datasets.
  • Social Impact Projects: Helping NGOs and researchers define problems and prepare clean datasets for AI-driven solutions.
  • AI Research & Experimentation: Allowing hobbyists, students, and developers to explore machine learning without deep technical knowledge.
  • Education & Learning: Providing students a hands-on way to learn about AI problem formulation, data preprocessing, and model training.

Acknowledgements

We would like to express our gratitude to our fellow team members for their valuable contributions:

Contributing

Contributions are welcome! Please open an issue or submit a pull request.

Contact

For any inquiries, reach out at neelkanthkhithani@gmail.com.

About

AI-Hobbyist is a platform that helps users generate problem statements, clean datasets, and train AI models using a no-code approach.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages