BrainBrew turns a fragmented semester into a reasoning system.
Students rarely fail because information is unavailable.
They fail because the semester is fragmented across:
- course pages
- assignments
- notes
- syllabi
- deadlines
- mental overhead
BrainBrew is the layer above that chaos. It helps a student understand:
- what matters now
- what is actually at risk
- which concept gaps are causing the risk
- what study action to take next
This is not a generic study chatbot. It is an academic workflow system built around reasoning, triage, and intervention.
Command Center • Assignment Recovery • Semester-Aware Chat • Study Mission Lab
- surfaces urgent work across the semester
- flags at-risk assignments
- shows readiness and remediation paths
- maps concepts across courses
- exposes weak areas before exams make them obvious
- connects assignment risk back to missing understanding
- generates quizzes, flashcards, and study guides
- grounds outputs in course context instead of generic prompts
- supports read-aloud and voice interaction
- searches universities
- pulls official university metadata
- imports transcripts
- matches coursework against official catalogs
Most academic AI tools stop at chat, summaries, or note cleanup.
BrainBrew goes further by reasoning over:
- the student’s current courses
- upcoming deadlines
- remediation state
- concept mastery
- imported academic history
The result is a product that behaves more like academic mission control than a single assistant window.
K2 Think V2
Used for:
- semester-aware chat
- study guide generation
- flashcard generation
- quiz generation
- concept extraction
- topic-level academic reasoning
Google Gemini
Used when K2 is unavailable during generation flows.
ElevenLabsfor primary text-to-speechKokoro-82Mas fallback TTS- backend speech transcription via
ElevenLabs
ReactTypeScriptViteZustandFramer Motion
PythonFastAPIUvicorn
SwiftSwiftUIAVFoundation
The current demo dataset is built around a realistic graduate computer science semester:
16:198:513Design and Analysis of Data Structures and Algorithms16:198:518Operating Systems Design16:198:533Natural Language Processing16:198:536Machine Learning
Rutgers is the current showcase context, but the product direction is broader: BrainBrew is designed as a general academic workflow layer for students navigating dense, multi-course semesters.
Live deployment:
- Web:
https://brain-brew.us - API:
https://api.brain-brew.us
flowchart LR
A[Student] --> B[BrainBrew Web or iOS]
B --> C[Command Center]
B --> D[Study Mission Lab]
B --> E[Knowledge Graph]
B --> F[Profile / Transcript Import]
C --> G[FastAPI Backend]
D --> G
E --> G
F --> G
G --> H[K2 Think V2]
G --> I[Gemini Fallback]
G --> J[ElevenLabs]
G --> K[Kokoro Fallback]
G --> L[Official University Sources]
G --> M[Demo Semester Data]
H --> N[Semester-Aware Reasoning]
I --> N
J --> O[Primary Voice Output]
K --> O
L --> P[University Search / Catalog Matching]
M --> Q[Courses / Assignments / Concepts]
Hosted on:
Render
Domain managed on:
Porkbun
Create backend/.env:
K2_API_KEY=your_k2_key_here
ENABLE_GEMINI_FALLBACK=true
GEMINI_API_KEY=your_gemini_key_here
ELEVENLABS_API_KEY=your_elevenlabs_key_here
ELEVENLABS_VOICE_ID=21m00Tcm4TlvDq8ikWAM
ELEVENLABS_TTS_MODEL=eleven_turbo_v2_5
KOKORO_VOICE=af_heart
KOKORO_LANG_CODE=a
KOKORO_SPEED=1.0
HF_TOKEN=your_huggingface_token_hereRun:
cd backend
pip install -r requirements.txt
python run.pyCreate frontend/.env if needed:
VITE_API_BASE_URL=http://localhost:8000/apiRun:
cd frontend
npm install
npm run devSee IOS_APP/README.md.
- Varesh Patel
- Aparajita Sarkar
- Sinchana S Arun
Built for HackPrinceton 2026.
BrainBrew turns academic chaos into a system that can reason, prioritize, and respond.




