Project Info
This project did not submit a demo video on Devpost.
Inspiration
In our journey as students, we have come to recognize the profound predicament that ensues when attempting to glean meaningful insights from lengthy video content. The struggle to synthesize information effectively from such resources is palpable. In light of this predicament, we embarked on a mission to craft a solution that not only ameliorates this issue but also bestows a boon to our fellow students by saving their precious time. We sought to empower them with a tool that not only facilitates the curation of their educational content but also ensures the utmost convenience by allowing them to revisit videos pertinent to specific topics.
What it does
Our innovative product, EduScribe, is endowed with multifaceted capabilities aimed at enhancing the educational experience of its users. It excels in the art of distilling high-quality summaries from YouTube videos, thereby rendering the extraction of essential insights an effortless endeavor. Furthermore, EduScribe takes the onus of maintaining a comprehensive record of the user's viewing history, facilitating easy reference to previously watched videos. This feature is invaluable for those seeking to revisit content related to a specific subject matter. In addition, EduScribe offers a succinct overview of search results based on the user's content history, thereby streamlining the process of finding pertinent information.
How we built it
The development of EduScribe was an endeavor that harnessed the capabilities of three cutting-edge technologies, namely Vectara AI, Together.ai, and Convex. The fusion of these technological marvels resulted in the creation of our robust product. Vectara AI contributed its prowess in frontend development, ensuring a seamless user interface. Together.ai played a pivotal role in facilitating data migration, a task often fraught with challenges. Convex, while a potent tool, presented its own set of challenges, particularly in comprehending the syntax for database querying. Additionally, the process of fine-tuning an AI model for summary generation and implementing zero-shot prompting for the LLaMa 2 models posed intricate challenges that we had to surmount.
Challenges we ran into
Our journey in developing EduScribe was not without its share of obstacles. Foremost among these was the formidable challenge of querying from the Convex database. The intricacies of database querying, coupled with the necessity of understanding and implementing the correct syntax, proved to be a daunting task. On the frontend, integrating Vectara AI presented its own set of difficulties, albeit surmountable. Data migration, while a crucial step in the development process, was not without its complications. Fine-tuning an AI model for summary generation, a task requiring precision and expertise, demanded a significant investment of time and effort. Moreover, implementing zero-shot prompting for the LLaMa 2 models, though a promising avenue, presented its own unique challenges that required creative problem-solving.
Accomplishments we're proud of
Our journey in developing EduScribe culminated in a series of accomplishments that we hold in high esteem. Most notably, we successfully crafted a full-stack project, reflecting our proficiency in both frontend and backend development. The crowning achievement of our endeavor was the creation of a sophisticated machine learning model for summary generation, a feat that showcases our commitment to innovation and excellence in the field of artificial intelligence.
What we learned
The development of EduScribe was a crucible of learning, where we immersed ourselves in a plethora of new and diverse technologies. Throughout this journey, we gained invaluable insights and knowledge, transcending our previous boundaries and expanding our technological horizons. Our acquisition of expertise in a variety of domains and the cultivation of our problem-solving skills have left an indelible mark on our development team.
What's next
The horizon of possibilities for EduScribe is broad and promising. Our future endeavors include the transformation of EduScribe into a browser extension, thereby increasing its accessibility and usability. This expansion will further enhance the educational experience of our users, making EduScribe an indispensable tool for the academic journey.
EduScribe
Using AI to turn watching minutes into moments
CalHacks 10.0 Project (Best Use of Vectara)
Inspiration
In our journey as students, we have come to recognize the profound predicament that ensues when attempting to glean meaningful insights from lengthy video content. The struggle to synthesize information effectively from such resources is palpable. In light of this predicament, we embarked on a mission to craft a solution that not only ameliorates this issue but also bestows a boon to our fellow students by saving their precious time. We sought to empower them with a tool that not only facilitates the curation of their educational content but also ensures the utmost convenience by allowing them to revisit videos pertinent to specific topics.
What it does
Our innovative product, EduScribe, is endowed with multifaceted capabilities aimed at enhancing the educational experience of its users. It excels in the art of distilling high-quality summaries from YouTube videos, thereby rendering the extraction of essential insights an effortless endeavor. Furthermore, EduScribe takes the onus of maintaining a comprehensive record of the user's viewing history, facilitating easy reference to previously watched videos. This feature is invaluable for those seeking to revisit content related to a specific subject matter. In addition, EduScribe offers a succinct overview of search results based on the user's content history, thereby streamlining the process of finding pertinent information.
How we built it
The development of EduScribe was an endeavor that harnessed the capabilities of three cutting-edge technologies, namely Vectara AI, Together.ai, and Convex. The fusion of these technological marvels resulted in the creation of our robust product. Vectara AI contributed its prowess in frontend development, ensuring a seamless user interface. Together.ai played a pivotal role in facilitating data migration, a task often fraught with challenges. Convex, while a potent tool, presented its own set of challenges, particularly in comprehending the syntax for database querying. Additionally, the process of fine-tuning an AI model for summary generation and implementing zero-shot prompting for the LLaMa 2 models posed intricate challenges that we had to surmount.
Challenges we ran into
Our journey in developing EduScribe was not without its share of obstacles. Foremost among these was the formidable challenge of querying from the Convex database. The intricacies of database querying, coupled with the necessity of understanding and implementing the correct syntax, proved to be a daunting task. On the frontend, integrating Vectara AI presented its own set of difficulties, albeit surmountable. Data migration, while a crucial step in the development process, was not without its complications. Fine-tuning an AI model for summary generation, a task requiring precision and expertise, demanded a significant investment of time and effort. Moreover, implementing zero-shot prompting for the LLaMa 2 models, though a promising avenue, presented its own unique challenges that required creative problem-solving.
Accomplishments that we're proud of
Our journey in developing EduScribe culminated in a series of accomplishments that we hold in high esteem. Most notably, we successfully crafted a full-stack project, reflecting our proficiency in both frontend and backend development. The crowning achievement of our endeavor was the creation of a sophisticated machine learning model for summary generation, a feat that showcases our commitment to innovation and excellence in the field of artificial intelligence.
What we learned
The development of EduScribe was a crucible of learning, where we immersed ourselves in a plethora of new and diverse technologies. Throughout this journey, we gained invaluable insights and knowledge, transcending our previous boundaries and expanding our technological horizons. Our acquisition of expertise in a variety of domains and the cultivation of our problem-solving skills have left an indelible mark on our development team.
What's next for EduScribe
The horizon of possibilities for EduScribe is broad and promising. Our future endeavors include the transformation of EduScribe into a browser extension, thereby increasing its accessibility and usability. This expansion will further enhance the educational experience of our users, making EduScribe an indispensable tool for the academic journey.
Analysis
View
Metric
- 12
- 1
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
- HTMLIn code
- JavaScriptIn code
- ReactIn code
- Tailwind CSSIn code
- TypeScriptIn code
- Node.jsClaimed
- PythonClaimed
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
146 KB
Source files
68
Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.
Repository
VinnyXP/EduScribe
81 files · 1.1 MB · @ 2404ebb
Structure
Interface
42 files · 52%Screens, components and styles rendered to the user.
Application logic
22 files · 27%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
- TypeScript86%
- JavaScript7%
- Markdown5%
- CSS1%
- HTML0%
Share of indexed source by file size. Binary and vendored files are excluded.
Dependencies
eduscribe/package.json
npm · 61- @clerk/clerk-react
- @hookform/resolvers
- @radix-ui/react-accordion
- @radix-ui/react-alert-dialog
- @radix-ui/react-aspect-ratio
- @radix-ui/react-avatar
- @radix-ui/react-checkbox
- @radix-ui/react-collapsible
- @radix-ui/react-context-menu
- @radix-ui/react-dialog
- @radix-ui/react-dropdown-menu
- @radix-ui/react-hover-card
- @radix-ui/react-icons
- @radix-ui/react-label
- @radix-ui/react-menubar
- @radix-ui/react-navigation-menu
- @radix-ui/react-popover
- @radix-ui/react-progress
- +43 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.
Feature verification
Full-stack app with React frontend + Convex backendVerified
Successfully crafted a full-stack project
Claimed on readmehigh confidenceeduscribe/src/App.tsx:13— React frontend calls the Convex action extractTranscript via useAction, wiring frontend to backendeduscribe/convex/myFunctions.ts:66— Convex backend actions/mutations implement the pipeline
YouTube transcript extractionVerified
Extracts and processes transcript text from a YouTube video URL
Claimed on readmehigh confidenceeduscribe/convex/myActions.ts:104— fetchTranscriptData action uses youtube-transcript package to fetch and merge transcript text
YouTube video summary generationVerified
Distills high-quality summaries from YouTube videos using an LLM
Claimed on readmehigh confidenceeduscribe/convex/myActions.ts:20— makeInferenceRequest posts a prompt to Together.ai's LLaMA-2 inference endpoint and returns generated texteduscribe/convex/myFunctions.ts:66— extractTranscript action fetches the transcript, formats it into a summary prompt via formatTranscript, and calls fetchAnalysis to produce the summary
Vectara-powered frontend/search integrationCode-supported
Vectara AI contributed to frontend development and search
Claimed on readmelow confidenceeduscribe/convex/myActions.ts:68— uploadToVectara action uploads the generated summary as a document to Vectara's upload API, but no querying/search of Vectara or any Vectara-driven frontend UI is found in src/
Viewing history storageCode-supported
Maintains a comprehensive record of the user's viewing history for later reference
Claimed on readmemedium confidenceeduscribe/convex/myFunctions.ts:38— addVideo mutation inserts video_url, transcript, and analysis into a yt_videos table, storing watched-video recordseduscribe/convex/myFunctions.ts:104— showAnalysis query, which would read back stored history, is a stub that just logs and returns the literal string 'beep'; the real retrieval logic is commented out, so history is never surfaced to the UI
Browser extension version of EduScribeClaimed only
Future plan to transform EduScribe into a browser extension
Claimed on readmemedium confidenceOverview of search results based on content historyClaimed only
Offers a succinct overview of search results based on the user's content history
Claimed on readmehigh confidence
An AI agent derived these features from the project’s Devpost page and readme, then searched the code for each one. Verified features are backed by cited code; claimed-only features had no supporting code, which is not by itself proof a feature is missing.
Export this project's context (description, README, evidence, key source files) to chat with an AI agent elsewhere.