Participants – Only for on-campus Technology students of Banasthali

Challenge 1: Bobble AI

Localised Content Generation using Generative AI
Welcome to the AI Challenge! Are you ready to push the boundaries of technology and create innovative solutions? Our challenge focuses on "Localised Content Generation using Generative AI."
Challenge Overview: Utilise the capabilities of Generative AI to create localised content that reflects the essence of different regions, including festivals. Take inspiration from the Bobble Keyboard for samples. The objective is to develop a solution that generates content with a local touch, incorporating elements such as clothing styles, backgrounds, text in local languages, and festival-specific content.
Key Requirements:
1. Localization Elements: Implement features that capture the local flavour, including but not limited to:
* Clothing styles representative of specific regions.
*Backgrounds that resonate with local aesthetics.
*Text and stickers in local languages or styles, e.g., a "thank you" sticker in the local language.
*Festival-specific content, incorporating elements associated with local celebrations.

2. Bobble AI Keyboard Exploration: Explore the Bobble Keyboard for inspiration and reference in creating diverse and culturally rich content. Analyse the samples to understand how localised and festival elements can enhance user experience.

3. Generative AI Implementation: Leverage Generative AI techniques to dynamically generate content that is not only contextually relevant but also resonates with the cultural nuances of different regions and includes festival-specific elements. Consider the efficiency of your solution by factoring in the size of the AI model and the inference time. Optimise the model for faster and more resource-efficient content generation

Submission Guidelines:
Submit a detailed explanation of your approach, highlighting the techniques and algorithms used in achieving localised and festival-specific content generation.
Provide a demo or prototype showcasing the application of your solution.
Emphasise user experience and the seamless integration of localised and festival elements.
Evaluation Criteria: Submissions will be evaluated based on the following criteria:
Localization Accuracy: The extent to which the generated content accurately reflects the cultural and regional elements, including festivals.
Festival Integration: Creativity and accuracy in incorporating festival-specific elements into the generated content.
Innovation: Uniqueness and creativity in implementing Generative AI for localised and festival-specific content generation.
User Experience: The intuitiveness and user-friendliness of the solution, considering the incorporation of localised and festival elements.
Technical Proficiency: The efficiency and effectiveness of the Generative AI techniques, including optimization of AI model size and inference time.
Get ready to embark on this exciting AI journey! We look forward to witnessing your creativity in Localised Content Generation using Generative AI, including the vibrant and diverse elements of local festivals. Best of luck!

Challenge 2: CoRover
Knowledge Base (Kb) Creation 
Create a basic question-answer system using Python and simple open-source technologies. The objective is to build a system where users input questions or text, and the system provides relevant answers based on predefined rules and matching keywords. 

Team should focus on implementing basic functionalities and user interactions within a short timeframe, fostering rapid prototyping and experimentation in the field of natural language processing.

Need To Have: 

Start with a small dataset containing sample questions and their corresponding answers. This dataset will serve as the initial knowledge base. No extensive preprocessing is required for this simplified version. 
Implement a straightforward keyword matching algorithm to match user input with predefined questions in the dataset. Use Python's string manipulation and comparison functions for basic matching. 
Design simple rules or patterns to determine the appropriate answer based on the user's input. For example, if the input question contains certain keywords, return a specific answer associated with those keywords. 
Try to create a simple command-line interface where users can input questions or text, and the system responds with relevant answers. Utilise Python's input() function for user interaction.

Nice To Have (Advanced):

Allow users to add new questions and answers to the system during runtime. 
Implement basic error handling mechanisms to handle unexpected user inputs gracefully. 
Design the system in a modular way so that additional features or functionalities can be easily integrated in the future.

Evaluation Criteria: 
The system should accurately respond to user queries based on the predefined rules and keyword matching. 
The user interface should be intuitive and easy to understand.
The system should handle various types of user inputs and provide appropriate responses or feedback. 

The advanced features listed under section C will also serve as the ranking basis for the finalists. 

Challenge 3: OpsTree Solutions

Create an MySQL Kubernetes operator

Challenge Overview:
You are tasked with creating a Kubernetes operator specifically tailored for managing MySQL instances. An operator is essentially a piece of software that automates the deployment, configuration, and management of applications on Kubernetes.

Key Requirements:

Stateful-Set with Persistent Volume: MySQL instances should be created using a StatefulSet, which is a way to deploy stateful applications on Kubernetes.
Persistent Volumes should be attached to ensure that data is stored persistently even if the MySQL pod is moved or restarted.
Configuration Management: Your operator should allow users to easily manage configurations for MySQL. This means users should be able to customize settings like database name, user credentials, etc., and your operator should apply these configurations to the MySQL instances.
Two Kubernetes Services: Cluster IP Service: A service with a Cluster IP, meaning it is accessible only within the Kubernetes cluster. This is useful for internal communication.
Headless Service: A headless service, meaning it doesn't allocate a Cluster IP. It is used for scenarios where each pod needs to be individually discoverable. In the context of MySQL, it can be helpful for scenarios like replication.
Logic for Updating MySQL StatefulSet/Configuration: The operator should include logic to update the MySQL StatefulSet and configurations dynamically. This could involve scaling the StatefulSet, rolling updates, or adjusting configurations without requiring manual intervention.

What You Need to Do:
1. MySQL StatefulSet:
   - Implement the creation of MySQL instances using a StatefulSet.
   - Ensure that Persistent Volumes are properly configured and attached to the MySQL pods.

2. Configuration Management:
   - Develop a mechanism for users to easily configure MySQL settings through the operator. This could be through custom resources or some other form of configuration.

3. Kubernetes Services:
   - Create two services for MySQL - one with a Cluster IP and another as a headless service.

4. Dynamic Updates:
   - Implement the logic for updating the MySQL StatefulSet and configurations when changes are made. This could involve watching for changes in custom resources or other triggers.

Why This Matters:
This challenge is about automating the deployment and management of MySQL on Kubernetes. By creating a specialized operator, you make it easier for developers and administrators to handle MySQL instances without diving deep into Kubernetes intricacies. It promotes scalability, flexibility, and ease of use in managing MySQL databases within a Kubernetes environment.

Challenge 4: Open Innovation Challenge

If you have any brilliant idea about solving a challenge that India could potentially use for innovation in any of the sectors such as EdTech, FinTech, AgriTech, E-Commerce, HealthTech, MedTech, BioTech, EnergyTech, CleanTech etc, you can build it using Artificial Intelligence.

Challenge 5: Bobble AI
Domain-Specific Conversational Chatbot using Large Language Models (LLM)

Overview:
The goal of this hackathon challenge is to design and implement a conversational chatbot using Large Language Models (LLM) tailored for a specific domain. Participants will develop a chatbot that effectively interacts with users within a predefined context, offering assistance, information, or engagement based on the domain chosen for the chatbot.

Example:
Mental Health: An assistant you can talk to when you are in stress, for relief.
AI Friend: A chatbot, which does conversation as a friend with the user.