Cognify: A reimagined reading experience solution

chevron-icon
Back
project-presentation-img
Presentation
Brainy Box
Project Owner

Cognify: A reimagined reading experience solution

Funding Requested

$75,000 USD

Expert Review
Star Filled Image Star Filled Image Star Filled Image Star Filled Image Star Filled Image 0
Community
Star Filled Image Star Filled Image Star Filled Image Star Filled Image Star Filled Image 4.5 (8)

Overview

Cognify is an innovative solution that combines artificial intelligence and a reinvented reading experience to transform the way people access knowledge. Over the recent years, we have observed, particularly among younger generations, a trend of deterioration in reading, concentration and comprehension skills. Research suggests that this decline may be related to digital habits, such as excessive time spent on social and entertainment media, which prioritizes quick engagement and doses of dopamine over deep, meaningful learning. With Cognify, we can harness the power of AI to reignite the passion for reading, making knowledge accessible and engaging again in our fast-paced digital age.

Proposal Description

How Our Project Will Contribute To The Growth Of The Decentralized AI Platform

Our project will help grow the AI platform by developing a highly specialized AI model, CognatorAI, that adapts and personalizes textual content for different users. 

By making the source code, datasets, and AI model open-source under the MIT license, we encourage collaboration and innovation.

Our model will also be available via API calls on SingularityNET's AI Marketplace, boosting usage and integration into various applications, leading to more engagement and development on the AI platform.

Our Team

Our team is exceptionally skilled and experienced, with extensive expertise in developing high-quality software solutions, robust infrastructure, and cutting-edge artificial intelligence. 

With a deep understanding of the latest technologies and industry best practices, we possess the comprehensive know-how to deliver top-tier solutions. This project has been meticulously designed to align with our delivery capabilities, guaranteeing a high-quality result.

 

View Team

AI services (New or Existing)

CognatorAI

Type

New AI service

Purpose

Adapt valuable knowledge by considering the user’s profile preferences interests and other individual characteristics through the specialized training of a Large Language Model (LLM) based on Llama 3-70B. Using fine-tuning techniques like LoRA and QLoRA on an original dataset this model will adapt any textual content (books articles documents...) to suit each user adjusting language complexity speed formatting and other aspects to ensure highly relevant results.

AI inputs

User profile preferences (interests and individualities) learning pace reading level interaction history and personalization settings (language complexity) and textual content data as raw text (books articles) and contextual info (summaries key points genre).

AI outputs

Personalized and adapted textual content including tailored versions of books articles and documents with language complexity and formatting adjusted to match the user's reading level and preferences ensuring users engagement accessible content suited to their individual needs and preferences.

Company Name (if applicable)

Brainy

The core problem we are aiming to solve

In Brazil, research indicates that the country has lost 4.6 million readers in the past four years, and only 16% of people have purchased books in the last 12 months. Meanwhile, the use of social media and online media has been increasing every year. 

Cognitive neuroscientists suggest that our digital habits, especially in younger generations, are "atrophying" our reading and comprehension abilities. This shift in media consumption, driven by the engagement and dopamine generated by social media, has significantly reduced access to knowledge and learning. 

This issue is particularly acute in Brazil and Latin America, where the decline in traditional reading habits exacerbates existing educational and informational disparities.

As more people turn to digital platforms for quick, engaging content, the depth and quality of knowledge acquisition suffer, leading to broader societal implications in terms of education, critical thinking, and informed citizenship.

Our specific solution to this problem

Cognify aims to help users grow in intellect, career, culture, and cognition.

To achieve this, we structured the project into two main areas:

AI (Artificial Intelligence) 
The first area aims to adapt valuable knowledge by considering the user’s profile, preferences, interests, and other individual characteristics through the effective and specialized training of a Large Language Model (LLM) based on Llama 3-70B. Fine-tuning techniques such as LoRA (Low-Rank Adaptation) and QLoRA (Quantized Low-Rank Adaptation) will be applied, using an original dataset of up to 200,000 contexts to ensure highly relevant results.

This model, named CognatorAI, will be capable of adapting any type of textual content (such as books, articles, documents, news, etc.) by utilizing various parameters from both the content and the user to adjust language, complexity, speed, formatting, and other aspects.

Experience
The second area aims to disseminate knowledge and enhance users' reading, cognition, and technical abilities through a reimagined, innovative, and modern reading experience.

A web platform and mobile-native app (Android and iOS), named Cognify, will be developed, giving users access to a vast library of content based on books, articles, and documents. This content will be adapted according to the user's profile and progress, gradually exposing them to more relevant, extensive, and complex material.

Project details

Cognify is an innovative solution that combines artificial intelligence and a reimagined reading experience to transform how people access knowledge.

In short, Cognify aims to help users develop their intellectual, professional, cultural and cognitive skills.

To achieve this, we structured the project into two main areas:

AI (Artificial Intelligence)
The first area aims to adapt valuable knowledge by considering the user’s profile, preferences, interests, and other individual characteristics through the effective and specialized training of a Large Language Model (LLM) based on Llama 3-70B. Fine-tuning techniques such as LoRA (Low-Rank Adaptation) and QLoRA (Quantized Low-Rank Adaptation) will be applied, using an original dataset of up to 200,000 contexts to ensure highly relevant results.

This model, named CognatorAI, will be capable of adapting any type of textual content (such as books, articles, documents, news, etc.) by utilizing various parameters from both the content and the user to adjust language, complexity, speed, formatting, and other aspects.

Use cases for CognatorAI

  • Learning

    • Students: Adapt textbooks, study materials, and exercises according to the student's knowledge level, learning pace, and style.

    • Professionals: Provide personalized summaries of scientific articles, technical reports, and documents relevant to each field.

  • Accessibility

    • People with cognitive difficulties: Simplify language, use short and direct sentences, and include visual aids to assist comprehension.

  • Reading

    • Classics adaptation: Make classic literary works more accessible to new readers by adjusting the language and contextualizing historical content.

  • Marketing

    • Marketing campaign: Adapt marketing campaign texts for a specific region or audience.

    • Blog personalization: Generate different versions of blog posts or articles customized for the user.

    • Advertisements: Adapt advertisement texts according to the target audience.

  • Entertainment

    • Personalized news: Adapt news content according to the user's interests, highlighting the most relevant topics and providing different levels of depth.

    • Interactive content creation: Generate games, quizzes, and other interactive activities based on texts, adapting difficulty levels and themes to the target audience.

  • Health

    • Healthcare awareness: Adapt informational materials about diseases, medications, and medical procedures for different audiences, from lay patients to healthcare professionals.

  • Law

    • Legal Document Simplification: Simplify laws, contracts, and other legal documents to facilitate understanding by the general population.

  • Finance

    • Financial Education: Create educational materials on investments, financial planning, and budget management in a personalized and accessible manner.

 

Experience
The second area aims to disseminate knowledge and enhance users' reading, cognition, and technical abilities through a reimagined, innovative, and modern reading experience.

A web platform and mobile-native app (Android and iOS), named Cognify, will be developed, giving users access to a vast library of content based on books, articles, and documents. This content will be adapted according to the user's profile and progress, gradually exposing them to more relevant, extensive, and complex material.

In the first interactions with Cognify, users will be asked questions to establish an initial profile and set up their interaction preferences. Users will provide information about their previous reading experiences, content consumption habits, personal preferences, and areas of interest. As users continue interacting with Cognify, their profiles will be automatically updated to reflect their evolving preferences and habits.

Note: Initially, Cognify will be adapted to Brazilian Portuguese. In the future we hope to make more languages ​​available thus covering a larger area of ​​knowledge dissemination.

To approach this area, we structured it into four categories: UI/UX, Content Dynamics, Gamification, and Connection.

UI/UX
In this category, Cognify provides an intuitive and accessible interface with reading and accessibility preference settings. Users can customize text size, fonts, and themes and access an audio mode with text-to-speech (TTS). 

It provides interactive tools and configurations such as quizzes, images, flashcards, curiosities, exercises, terms, and concepts.

Content Dynamics
In this category, Cognify offers various ways to consume content, including interactive and dynamic methods such as summaries, quizzes, flashcards, terms and concepts, curiosities, exercises, notes, practical applications, lessons, and more.

All these interactive and dynamic contents are generated and adapted by the CognatorAI for each user, respecting their individualities.

Gamification
In this category, Cognify features several gamified elements to increase user engagement, retention, and motivation. Users encounter checkpoints while consuming content, earning progress points, new dynamic content, and interactions upon reaching each one. Additional features include daily challenges, achievements, leaderboards, and competitions among users.

Progress and interaction/use points are used as metrics to evaluate content engagement and for continuous training of CognatorAI.

Connection
In this category, Cognify present various connectivity elements typical of social and media networks. Interaction elements include liking specific content or excerpts, participating in discussions and debates, and sharing posts on other platforms, promoting a dialogue-rich environment.

These resources help create an engaged and interactive community where connections are strengthened through shared and collaborative experiences.

All these interactions are also used as metrics for continuous training of CognatorAI.

Use cases for Cognify

  • Casual reader seeking personal development

    • User Need 

      • Carlos is a casual reader who wants to develop personally but feels unmotivated by long and complex self-help books.

    • Solution Approach 

      • User Profile: The platform identifies Carlos' interests in personal development, such as time management, productivity, and emotional intelligence.

      • Initial Content: The AI provides short summaries based on books and introductory articles on these topics, using accessible and engaging language.

      • Continuous Motivation: As Carlos progresses, the platform presents deeper and more complex challenges gradually and motivatingly.

      • Dynamic Progress: Carlos receives reading recommendations based on his progress and feedback, with suggestions for books, articles, and videos that complement his learning.

      • Interactivity: Throughout the journey, the platform offers practical exercises and reflections for Carlos to apply personal development concepts in his daily life.

 

  • Software developer with limited time for technical updates

    • User Need

      • Rafael is a software developer with a busy schedule who wants to improve his technical knowledge but lacks time to search for and read articles, documents, and books on the subject.

    • Solution Approach 

      • User Profile: The platform identifies Rafael's technical interests, such as new programming languages, emerging frameworks, DevOps practices, and agile methodologies.

      • Content Customization: The platform personalizes content according to Rafael's interests and knowledge level, ensuring he receives only relevant and up-to-date information.

      • Summaries and Highlights: Long articles and technical documents are condensed into easy-to-read summaries with key points highlighted, saving time and facilitating information assimilation.

      • Dynamic Recommendations: Based on Rafael's learning history and feedback, the platform recommends daily content, adjusting to his specific needs and interests.

      • Interaction and Practical Application: Rafael has access to quick quizzes and practical exercises to test his knowledge and immediately apply new concepts, as well as forums where he can interact with other developers and experts.

      • Text-to-Speech: To maximize time use, the platform offers a text-to-speech feature, allowing Rafael to listen to articles and tutorials while performing other daily tasks.

 

  • Teenager spending too much time on social media and finds books boring

    • User Need

      • Lucas is a 16-year-old teenager who spends most of his day on social media and finds books boring and uninteresting. His parents are concerned about his lack of interest in reading and want him to develop healthier knowledge consumption habits.

    • Solution Approach 

      • User Profile: The platform identifies Lucas' interests based on his favorite activities on social media, such as games, sports, and technology.

      • Gamification: The platform adds gamification elements to content, like checkpoints, achievements, and leaderboards. Lucas earns points by completing readings and challenges, unlocking new content, competing with friends, and earning achievement badges.

      • Connection: The platform lets Lucas like content, join discussions, and share excerpts with friends, creating a social and collaborative learning environment. He can also follow authors and comment on reviews and articles, promoting active dialogue.