WorkwiseAI

PARIN ACHARYA

Generative AI Hackathon with IBM Granite

Quick Links

• Overview
• Features
• Technology Used
• Getting Started
  • Project Structure
  • Prerequisites
  • Installation
  • Usage
  • Testing
• Live URL Deployment
• Roadmap of Project Tasks
• Contributing
• License
• Acknowledgments

Hackathon page

---

Overview

WorkwiseAI is an AI Agent framework that pinpoints business workflow inefficiencies and streamlines processes using cutting-edge IBM Granite models.

Some key use cases of IBM Cloud Services in enhancing business workflows

1. Vodafone, a global communications leader, is using IBM Watson to simulate and analyze digital disucssions with its AI powered virtual agent, reducing testing timelines to under 1 minute. Read more
2. Artefact, a leading French Bank uses a portfolio of personas represented by AI identities, allowing professionals to reveal crucial insights from customer behavior. Read more

---

Features

• Business workflow context grounding using a knowledge base of company documents, in IBM Vector Indexes.

• Cutting-edge AI Insights in generative tasks through the state-of-the-art granite-3-8b-instruct foundation model. See Model card

• Workflow visualization dashboard in Streamlit with clear, interactive widgets to represent large data.

• WorkwiseAI 'S3' Agent - a ground-up agent pipeline built on top of the LangChain Framework in IBM AgentLab to analyze business workflow inefficiencies. This enables powerful support for 3 worker agents defined as follows.

  • Summarizer agent: Breaks down business workflow steps from the context of the knowledge base to reveal critical insights and inefficiency factors.
  • Scorer agent: Evaluates each step in the current workflow using a custom scoring model alongside industry-relevant metrics.
  • Suggester agent: Access to tools like WikipediaQuery, GoDuckGoSearch, and RAGQuery to suggest actionable improvements to each step in the workflow.

  

  

• Chatbot User Interface grounded in business context documents and offers multi-language support and advanced Natural Language Processing.

  

---

Getting Started

Project Structure

└── workflow-agent/
    ├── .images
    ├── LICENSE
    ├── README.md
    ├── main.py
    ├── react-agent.py
    ├── requirements.txt   
    ├── test_data.pdf
    ├── utils.py
    └── utils.py

Prerequisites

Before getting started with workflow-agent, ensure your runtime environment meets the following requirements:

• Programming Language: Python
• Package Manager: Pip

Installation

Install workflow-agent using one of the following methods:

Build from source:

1. Clone the workflow-agent repository:

❯ git clone https://github.com/ParinAcharyaGit/workflow-agent

2. Navigate to the project directory:

❯ cd workflow-agent

3. Install the project dependencies:

Using pip  

❯ pip install -r requirements.txt

Usage

Run workflow-agent using the following command: Using pip  

❯ python {entrypoint}

Testing

Run the test suite using the following command: Using pip  

❯ pytest

---

Live URL Deployment

You can access the live deployment here Visit Live Deployment

---

Roadmap of Project Tasks

Completed

• Task 1: Implement vector index endpoints and implement document upload feature.
• Task 2: Implement S3 agent worflow in IBM AgentLab
• Task 3: Enable key side-by-side vizualizations using Streamlit Widgets for comparison with S3 agent suggestions
• Task 4: Prototype Documentation and Demo

In progress

• Task 5: Firestore Database and User Authentication
• Task 6: Complete user testing and agent pipeline testing
• Task 7: Expand tool support for S3 Agent
• Task 8: Advanced NLP and vector embedding techniques for user chatbot conversations

---

Contributing

• 💬 Join the Discussions: Share your insights, provide feedback, or ask questions. Feel free to reach out on LinkedIn!
• 🐛 Report Issues: Submit bugs found or log feature requests for the workflow-agent project.
• 💡 Submit Pull Requests: Review open PRs, and submit your own PRs.

Contributing Guidelines

1. Fork the Repository: Start by forking the project repository to your github account.
2. Clone Locally: Clone the forked repository to your local machine using a git client.

   git clone https://github.com/ParinAcharyaGit/workflow-agent

3. Create a New Branch: Always work on a new branch, giving it a descriptive name.

   git checkout -b new-feature-x

4. Make Your Changes: Develop and test your changes locally.
5. Commit Your Changes: Commit with a clear message describing your updates.

   git commit -m 'Implemented new feature x.'

6. Push to github: Push the changes to your forked repository.

   git push origin new-feature-x

7. Submit a Pull Request: Create a PR against the original project repository. Clearly describe the changes and their motivations.
8. Review: Once your PR is reviewed and approved, it will be merged into the main branch. Congratulations on your contribution!

Contributor Graph

---

License

For more details, refer to the LICENSE file.

---

Acknowledgments

Resources, guides and support

1. IBM Documentation
2. Lablab.ai Discord Channel