Livin: seeing is understanding

The front page, this is what the user first sees
A place to allow the user to voice their questions
A loading screen, while we look through the databases and talk to the agents
A collection of videos relating to the query along with the answer to the query
A query search to see all the videos where a certain keyword was detected
Making a plan of actions to take that can help the user achieve their goals
A plan section to tell the user actions to take to get their desired outcome

What It Does

Unified Data Aggregation: Collects your daily events, activities, and interactions mainly from video but from various data sources.
Intelligent Curation: Uses advanced AI (Gemini 2.0 Flash) to analyze patterns and highlight key moments using video data that was uploaded Aperature DB.
Automated Task Management: Leverages AI agents (Agno) to handle routine organization, letting you focus on life’s big picture. Used Agno to make a web search AI agent to search for products.
Dynamic Visualization: Creates interactive timelines and summaries that present your personal journey in a clear, engaging way.

Personal Growth: Identify trends in your habits and well-being to drive positive lifestyle changes. Also help you relieve pain and save doctors visits by using video data along with gemini and agno AI agent to recommend solutions to problems.
Memory Enhancement: Revisit significant moments and milestones with intelligent, easily navigable summaries.
Efficiency Boost: Automate everyday tasks like scheduling and data organization to free up more time.
Digital Legacy: Build a curated archive of your experiences that you can reflect on or share with loved ones.

AI Agents: Developed specialized agents to continuously capture and analyze your daily data.
Knowledge Graph Database: Employed Apeture DB to store and interconnect diverse data types, ensuring efficient retrieval.
Retrieval-Augmented Generation (RAG): Transformed raw data into meaningful insights and actionable summaries. By using Aperture DB and Gemini 2-0 Flash.
User-Friendly Frontend: Designed an intuitive interface that lets you interact with and personalize your life timeline.
*Google Cloud Speech-To-Text: * Used Google Cloud Speech-To-Text to get the audio transcribed to text which was later used as the text for the query.

Data Integration: Merging information from various devices and formats into one cohesive system was complex.
System Scalability: Handling the continuous influx of high-volume data required robust infrastructure and frequent optimization.
Agent Coordination: Ensuring multiple AI agents worked in harmony without redundancy posed significant challenges.
Privacy & Security: Protecting sensitive personal data while providing rich insights demanded strict adherence to privacy standards.

Holistic Life Overview: Successfully created a system that provides a comprehensive view of your daily life, highlighting both routine and significant events.
High-Quality Summaries: Developed AI-driven methods that automatically identify and present key moments with minimal manual effort.
Streamlined Automation: Integrated AI agents that significantly reduce the grunt work of managing personal data.

Everyday Data is Valuable: Routine data, when analyzed correctly, can reveal deep insights about your habits and well-being.
User-Centric Design is Crucial: An intuitive, non-intrusive interface is key to seamlessly integrating LifeOS into your daily routine.
Collaboration Enhances Outcomes: Coordinating multiple AI components leads to more coherent and meaningful insights.
Privacy Must Come First: Handling personal data responsibly is essential for building trust and ensuring long-term success.

Enhanced Personalization: Develop adaptive models that learn from your unique lifestyle to offer even more tailored insights.
Expanded Data Sources: Integrate additional smart devices and platforms for a richer, more complete picture of your life.
Collaborative Sharing: Enable features for sharing select parts of your timeline with friends and family.
Continuous Improvement: Refine AI models and agent coordination to further reduce manual oversight and boost accuracy.
Automate purchases with authentication: Use Stych to get authentication to allow the agents to go ahead and purchase the products.

I worked on the backend, to get the audio recorded, and then stored into a file as a .wav format and then transcribed the audio, as a prompt for gemini. I achieved this using python, Speech-To-Text from Google, and also used the Gemini API. I also used agno and browser use to get the agent to make google searches for you and get your product links.

Sujan Nandikol
likethemonth Lee
leeu3581 lee
Terell Brown
I've been missing out on the tech culture for too long. Time to soak it up in all it's glory.

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.