Team formation and Project Proposal #

Link to repository #

https://github.com/IU-Capstone-Project-2025/Ego_AI

Team members #

Brainshtorming #

1. AI-Powered Study Assistant

• Idea: An AI service to help students prepare for exams by providing personalized learning materials.
• MVP Features:
  • Personalized quiz/test generation
  • User authentication
  • Subject-specific curated study materials (initially for one subject)

2. AI Productivity Agent (Web-Based)

• Idea: An AI assistant that helps with problem-solving, time management, and routine optimization.
• MVP Features:
  • Calendar integration
  • Task analysis & prioritization
  • Dashboard with progress tracking
  • Personalized content filtering

3. AI-Generated Visual Novel

• Idea: An interactive AI-driven visual novel where users influence the story.
• MVP Features:
  • Text generation based on initial parameters (genre, scene setup)
  • Branching storylines with player choices

4. AI-Powered Travel Guide

• Idea: A smart travel assistant that recommends places based on user interests and real-time data.
• MVP Features:
  • User profile & preference settings
  • Integration with maps/weather APIs
  • Personalized location summaries & recommendations

Project name: Ego AI #

Project Idea #

The objective of the project is to create an AI-powered assistant that learns from the user’s schedule and enhances its intelligence over time by adapting to their habits and activities. The assistant will be able to suggest activities, suggest suitable times for activities, and provide personalized recommendations based on the user’s preferences and interests.

Target users #

The target users of the project are students, professsors, and anyone who wants to improve their productivity and efficiency. The assistant will be designed to be user-friendly and accessible to individuals of all ages and backgrounds.

User stories #

Used Abbreviations: BR - business requirements, UR - user requirements

1. I want to have a clear and adjustable schedule (BR)

   1.1. As a user, I want the AI-assistant to show me my tasks and offer refinement, so I can achieve mo**Could implement:**re goals. (UR)

   1.2. As a user, I want the AI-assistant to provide an overall-analytics (dashboard) for my time-table, so I can have a vision of all the activities and achieve work/life balance. (UR)

2. I want to be able to communicate my ideas and save my time (BR)

   2.1. As a user, I want the AI-assistant to implement a speech-to-text recognition, so I can quickly add tasks or brainstorm. (UR)

   2.2. As a user, I want the AI-assistant to summarize the documents/long texts, so I can save my time reading them. (UR)

3. I want the personalized help and recommendations (BR)

   3.1. As a user, I want the AI-assistant to form a model of my interests and compile it to a knowledge base, so I can address it when searching activities/news to explore. (UR)

   3.2. As a user, I want the AI-assistant to analyze queries in natural language efficiently, so I can lead dialogs, compilations effortlessly. (UR)

   3.3. As a user, I want the AI-assistant to support some geo-integration and offer me places to visit, so I can diversify my leisure time. (UR)

4. I want to ensure the safety and affordability of the product I use (BR)

   4.1. As a user, I want the AI-assistant to provide data-safety agreement and offer high-quality protection of connected accounts, so I can feel that my personal info is protected. (UR)

   4.2. As a user, I want the AI-assistant be reasonably priced, so it is worth the money I spend. (UR)

Marketing analysis #

✍️ Reclaim.AI:

A calendar helper with integrated AI, that creates a centralized vision (with dashboards) on your daily schedule.

Pricing:

• $10/month

Most common features:

• Integration with Google
• priorities tracker
• smart time-blocking feature to readjust

Outstanding:

• amazing dashboards structure
• smart meetings
• habits incorporation
• time-blocking

Could implement:

• inspired by their stats, smart time-block and google calendar integration
• overall interface is intuitive and clean

✍️ Sider

A chrome extension as all-in-one AI-tools option

Prising:

• basic - $8.33

• unlimited - $25

• pro - $16,6/month

Most common features:

A variety of smart chat-bots/AI-tools in one place

Outstanding:

• lots of services combined
• wisebase
• turn info into summaries and generate drafts/FAQs from them

Could implement:

• integrate some chat options along with wisebase idea to make a natural conversation flow experience

✍️ Akiflow

A project management tool with AI-support, that enhances experience in time-management

Pricing:

• $34/month

• $19/yearly

Most common features:

• a unique task tracking app with useful productivity tracking
• AI-workflows is an interesting feature

Outstanding:

• co-pilot learns habits
• works offline (an app)
• ai-workflows

Could implement:

• ai-workflows looks like something we can build upon

✍️ Clockwise

AI Time Management System to Create the Perfect Schedule

Pricing:

• $7.75/month

Most common features:

• Analyze complex schedules
• Reorganize events
• Find time in groups
• Book meetings

Outstanding:

• Processing huge amounts of data
• Focus blocks

Could implement:

• Dynamic rescheduling

✍️ SkedPal

An intelligent time and task management app that combines calendar and to-do list with AI-powered schedule optimization

Pricing:

• $14.95/month

Most common features:

• Intelligent task planning based on priorities
• Time Maps (time zone visualization)
• Dynamic rescheduling
• Kanban board for priorities
• Nested tasks and notes
• Integration with Google Calendar, Outlook, Asana, Zapier

Outstanding:

• Unique Time Maps and Budgets system for time control by categories
• Automatic rescheduling when tasks are missed
• Relative ranking of priorities, not rigid categories
• Time scaling by selecting a frame (new calendar design)
• Support for consistent task scheduling even with unplanned elements
• Support for complex projects with nested tasks

Could implement:

• Implementation of Kanban board with WIP limits to optimize task flow
• Intelligent prioritization and visual task management
• Dynamic rescheduling with AI
• Integration with corporate calendars and task managers
• Visualization and analytics of user productivity

🌎 Mapbox Navigation SDK

Navigation with AI assistant MapGPT

Pricing:

• Usage-based pricing $0

Most common features:

• Voice control
• Range prediction
• 3D visualization

Outstanding:

• Natural voice commands
• 2075+ attractions
• Contextual queries

Could implement:

• Voice control
• Geo-visualization

🌎 Yandex.Maps (API) Geoinformation services with interactive and static maps, routing, organization search

Pricing:

• Free up to 1000 DAU
• from 195,000 ₽/year for 5000 DAU
• rices increase with increasing DAU

Most common features:

• Route building
• Objects on the map via ObjectManager
• Integration with Yandex.Tables and DataLens
• Navigation, traffic jams, transport schedules

Outstanding:

• Personalization and AI ranking algorithms
• Analysis of user interactions with company cards

Could implement:

• Implementation of AI recommendations for personalized selection of places and routes
• Use of neural networks for filtering and improving data quality
• Integration of voice search and AI prompts
• Development of interactive 3D maps with detailed infrastructure
• Analytics of user behavior for optimization of offers and business proposals

🌎 Google maps (API) Global mapping service with navigation, POI search, routing, 3D visualization, Street View and integration with other Google services

Pricing: the cost of Google Mapd API depends on many factors and is divided into several levels (Essentials, Pro, Enterprise), there is no static price, it is better to conduct a separate analysis and consider on the website https://mapsplatform.google.com/pricing/

but up to some point the cost is $0 - Search and recommend places

Most common features:

• Route building and traffic optimization
• Street View and 3D visualization
• Track prices for hotels and flights
• Integration with Google Assistant

Outstanding:

• Gemini Generative AI for creating personalized routes and recommendations
• Analysis of user screenshots to automatically create lists of places (AI Overviews)
• Real-time traffic and traffic forecasting
• Price tracking and discount notifications
• Interactive AI responses to queries using personal context (Gmail, calendar)
• AI Mode with Deep Search and Visual Recognition via Camera

Could implement:

• Implementing AI for personalized recommendations and travel planning
• Automating route creation based on user data and preferences
• Integrating voice and visual AI search
• Using AI to analyze and improve data quality
• Developing features for price tracking and notifications

Project timeline #

• Week 1: Establish the project’s boundaries, identify the target audience, select the technical stack, and set a timeline.

• Week 2: Begin designing the project using Figma, and commence development of the backend and frontend.

• Week 3: Present the Minimum Viable Product (MVP), focusing on implementing essential features.

• Weeks 4-6: Conduct testing, address bugs, and enhance the project’s functionality, add new functions.

• Week 7: Launch the initial version of EGO AI.

Choosen tech stacks #

• Backend: Python, Python FastAPI, PostgreSQL

• Frontend: React, React Native

• Design: Figma

• DevOps: docker, docker-compose

• ML: Hagging Face API, Python, PyTorch, Scikit-learn

Weekly commitments #

Individual contribution of each participant #

• Dmitriy Ryazanov: part of marketing analysis, Tech stacks discuss

• Diana Tsoi: Create user stories, part of the marketing analysis

• Stepan Sarantsev: Tech stacks discuss

• Andrey Zhdanov: Tech stacks discuss

• Artyom Lapin: Helped with Docker containers, docker-compose, Tech stacks discuss

• Mikhail Sofin:Tech stacks discuss

• Salavat Zaynulin: write report, Team organization, creating project repository setup,Tech stacks discuss

Confirmation of the code’s operability #

We confirm that the code in the main branch:

• In the working condition

• Run via docker-compose