docXmart is your intelligent multilingual assistant for all document-related tasks, including summarizing, extracting information, translating, and more. It leverages the AI71 API to provide accurate and detailed responses based on user inputs.
- Document Upload: Supports .txt, .md, .pdf, and .docx formats.
- Language Detection: Automatically detects the language of the uploaded document.
- Task Selection: Offers a variety of tasks such as summarizing, extracting information, translating, and more.
- Multilingual Translation: Supports translation into multiple languages.
- Real-time Response Generation: Generates responses in real-time with smooth flow and transition.
-
Upload Your Document:
- Provide a document in .txt, .md, .pdf, or .docx format.
-
Describe Your Request:
- Choose a category from the prompt list:
- General Prompts:
- Translate the document to another language
- Review and correct grammatical errors in the document
- Generate a concise summary of the document
- Extract specific information such as dates, names, and places from the document
- Determine the sentiment (positive, negative, neutral) expressed in the document
- Create new text based on the document’s content
- Expand on ideas or topics mentioned in the document
- Identify the main theme or subject of the document
- Rewrite sections of text to improve clarity or readability
- Identify and classify entities such as people, organizations, locations, and dates in the document
- Suggest improvements for style and readability of the document
- Identify and list important keywords or phrases from the document
- Analyze and identify recurring themes or topics within the document
- Educational Prompts:
- Generate text for PowerPoint slides based on the document
- Generate study notes based on the document
- Summarize educational content for easier understanding
- Create quiz questions based on the educational document
- Develop lesson plans or teaching materials from the document
- Provide examples and analogies to explain difficult concepts
- Identify and suggest additional resources for further reading
- Generate discussion questions to encourage critical thinking
- Analyze the document for educational standards alignment
- Suggest some possible visual aids that could be created based on the document
- Assess the readability level of the educational content
- Legal Prompts:
- Summarize legal documents and extract key points
- Translate legal documents to another language
- Identify key legal terms and definitions within the document
- Check for compliance with legal standards and regulations
- Draft legal contracts or agreements based on provided information
- Analyze and identify potential legal risks or issues
- Provide legal citations and references for document content
- Review and correct legal document formatting and structure
- Generate a timeline of events based on the legal document
- Assess the strength of arguments in legal documents
- Review and improve a resume
- Summarize a CV
- Extract key skills and qualifications from a resume
- Generate a professional summary based on a resume
- Tailor a resume for a specific job
- Identify strengths and weaknesses in a resume
- Convert a resume to a different format
- Review a resume for ATS (Applicant Tracking System) optimization
- Create a cover letter based on a resume
- General Prompts:
- Choose a category from the prompt list:
-
Submit Your Request:
- Click 'Submit' to process your document.
-
Review the Response:
- docXmart will analyze your document and generate a detailed response based on your instructions.
- Clone the repository:
git clone https://github.com/yourusername/docXmart.git
- Install the required dependencies:
pip install -r requirements.txt
- Run the Streamlit app:
streamlit run app.py
- Navigate to the local server where the Streamlit app is running.
- Upload a document using the file uploader.
- Select the desired task from the prompt list.
- Click 'Submit' and wait for the response to be generated and displayed.
- User Uploads Document: The user uploads a document using the file uploader.
- Document Processing: The document is read and processed to detect its language and split into manageable chunks if necessary.
- Task Selection: The user selects a task and, if applicable, a target language for translation.
- Prompt Generation: Prompts are generated based on the user's task selection and document chunks.
- Response Generation: The AI71 API is called to generate responses for each document chunk.
- Display Response: The responses are displayed in real-time with a smooth flow to the user.
Below is a high-level flow for the docXmart project.
6C74
- Streamlit
- ai71
- chardet
- fitz (PyMuPDF)
- langid
- pycountry
- re
- asyncio
We welcome contributions to docXmart! Please fork the repository and submit pull requests.
This project is licensed under the MIT License.
We thank the developers of the AI71 API and all the open-source contributors especially Malaika Farooq, Uroosa Saad, Fareed Khan, and Usama Nisar Khan who made this project possible.