Project Info
Inspiration
Remembering things is hard – and staying on top of your self improvement and reflection while juggling a busy schedule is even harder. As a team of busy students, one of whom recently started working a full time job, we decided to build Mirror as a tool to leverage unique smart glasses technology to revolutionize the way people handle self-improvement and reflection.
What it does
Mirror is an AI-powered app compatible on all AugmentOS-enabled smart glasses which transcribes your conversations and generates actionable insights and summaries tuned for reflection in real time. Our processing agents will automatically update on the Mirror web interface as you speak, intelligently picking out the most appropriate points to give a well-rounded overview of your day, extracting todos, a general snapshot of your day, and important reflection questions. Just as important as understanding the specific details of your day is being able to look back in time at past memories and see how they fit into the context of the rest of your memories. Mirror offers full-text semantic search through all memory content to allow for quick and easy retrieval and recall of specific details even if you only have a small idea of what you’re looking for. Furthermore, visualizing the relationships between memories is made easy and beautiful via Mirror’s 3-Dimensional memory map which displays latent connections between memories. Each edge and connected component tells a story and provides insights into the relationships between your experiences. The best part – all of this is generated via our processing pipeline in real time as you talk! Throughout the entire day, Mirror will always be up to date with your latest memory insights.
How we built it
Mirror is built in 3 parts: The AugmentOS Smart Glasses Client is an Android application written in Java which handles the interface with the smart glasses – handling both the voice transcription and relaying text back to the user via the display. Shoutout to the Mentra team for providing mentorship and lending us the Even Realities G1 Glasses that we built Mirror to work with. The backend is written in TypeScript via NextJS and handles interactions with both the smart glasses client and the web interface. Using ElasticSearch, Mistral AI, Vercel’s NextJS, and Codeium’s Windsurf, we created an intuitive interface to interact with your thoughts throughout the day.
What we learned
We learned that working with experimental hardware is hard – and persistence is key. We are the R&D for the software for these glasses, testing the limits of what’s possible with them and trying to push the field forward.
What's next
We would love to continue expanding Mirror’s interaction with our users, especially in terms of active reflection throughout the day. We envision a world where our agents can prompt you during conversations about related past conversations, remind you of your past reflections, and help guide you towards a more productive, social future.
Mirror
Reflecting on your past work, relationships, and learning has never been easier. Mirror turns your minute-to-minute into a comprehensive and indexable knowledge base in the form of a journal.
Mirror is an AI-powered app compatible on all AugmentOS-enabled smart glasses which transcribes your conversations and generates actionable insights and summaries tuned for reflection in real time. Our processing agents will automatically update on the Mirror web interface as you speak, intelligently picking out the most appropriate points to give a well-rounded overview of your day, extracting todos, a general snapshot of your day, and important reflection questions.
by: Sarah Su, Claire Wang, Kevin Zhu, Sanjith Udupa
Mirror is built in 3 parts:
Taken from our devpost
The AugmentOS Smart Glasses Client is an Android application written in Java which handles the interface with the smart glasses – handling both the voice transcription and relaying text back to the user via the display. Shoutout to the Mentra team for providing mentorship and lending us the Even Realities G1 Glasses that we built Mirror to work with.
The backend is written in TypeScript via NextJS and handles interactions with both the smart glasses client and the web interface. Using ElasticSearch, Mistral AI, Vercel’s NextJS, and Codeium’s Windsurf, we created an intuitive interface to interact with your thoughts throughout the day.
Furthermore, visualizing the relationships between memories is made easy and beautiful via Mirror’s 3-Dimensional memory map which displays latent connections between memories. Each edge and connected component tells a story and provides insights into the relationships between your experiences.
The best part – all of this is generated via our processing pipeline in real time as you talk! Throughout the entire day, Mirror will always be up to date with your latest memory insights.
Watch our video:
Link to Android app repo.
You can run the app yourself by using Android Studio (and combining the AugmentOS/Core app on an android phone with any smart glasses that has a mic) and sending a Pinggy tunnel to our web server.
Analysis
View
Metric
- 15
- 14
- 7
- 4
Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.
Technology
- CSSIn code
- Next.jsIn code
- OpenAIIn code
- ReactIn code
- Tailwind CSSIn code
- TypeScriptIn code
- JavaClaimed
- Mistral AIClaimed
6 of 8 appear in the indexed code. 2 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.
AI coding agents
No AI coding agent signals were found in this repository.
Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.
Codebase size
Source size
54 KB
Source files
31
Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.
Repository
kevinzhu12/treehacks-glasses
51 files · 3.6 MB · @ 5dd5b9f
Structure
Interface
15 files · 29%Screens, components and styles rendered to the user.
API & routing
7 files · 14%Request entry points: routes, handlers and controllers.
Application logic
10 files · 20%Domain rules, services and shared utilities.
Supporting
Layers are inferred from where files sit in the tree, not from reading the code. A project that names its directories unconventionally will read oddly here — open the file browser to check anything the diagram implies.
Languages
- TypeScript91%
- Markdown5%
- CSS4%
Share of indexed source by file size. Binary and vendored files are excluded.
Dependencies
package.json
npm · 24- @elastic/elasticsearch
- @mistralai/mistralai
- aframe
- dotenv
- fetch
- fs
- next
- openai
- react
- react-dom
- react-force-graph
- react-force-graph-3d
- react-markdown
- remark-breaks
- +10 more
Declared in the repository’s manifests at the indexed commit. A declared package is not proof it is used, and runtime dependencies are listed first.
This project’s features have not been analysed yet.
Export this project's context (description, README, evidence, key source files) to chat with an AI agent elsewhere.