Prompt Prep | Precision Study Ecosystem

---

Prompt Prep is a custom-engineered full-stack platform designed to synthesize vast amounts of educational data into high-utility learning assets. By integrating vector-based similarity search with state-of-the-art language models, the system enables users to parse documents deeply, generate rigorous academic evaluations, and interact with their notes through a precisely grounded conversational layer.

Core Capabilities

• Semantic Data Harmonization: Multi-format document parsing (PDF, MD, TXT) with intelligent chunking tailored for high-precision retrieval.
• Grounded Conversational Intelligence: A retrieval-augmented query layer that guarantees accuracy by sourcing responses directly from private datasets.
• Automated Knowledge Evaluation: Dynamic synthesis of comprehensive MCQ sets with automated scoring and exhaustive reasoning explanations.
• Rapid-Recall Synthesis: Algorithmic extraction of core concepts into structured flashcards for accelerated knowledge retention.
• Adaptive LLM Orchestration: Dual-engine intelligence spanning Google Gemini and Groq, featuring real-time provider switching and resilience.

Technical Architecture

The architecture leverages a high-concurrency Node.js environment coupled with a vectorized relational database to maintain low-latency response times during complex RAG operations.

graph TD
    UI[Interactive React Interface]
    Core[Node.js Orchestration layer]
    Logic[Service Architecture]
    Storage[(Supabase Hub)]
    Neural[Gemini & Groq Models]

    UI -->|API Requests| Core
    Core --> Logic
    Logic -->|Vector Indexing| Storage
    Logic -->|Semantic Context| Neural

Loading

Engineered Architecture Patterns

• Provider Strategy: Decoupled AI interfaces allowing seamless transitions between different LLM backends.
• Factory Orchestration: Centralized management for content generation and document processing flows.
• Entity Repository: Robust data persistence model powered by Prisma ORM.

---

Implementation Foundations

Layer	Environment	Purpose
Logic Engine	Node.js / TypeScript	High-concurrency operations and type-safety
UI Framework	React 18 / Framer	Fluid interactions and state-driven design
Data Hub	Supabase (Postgres)	Relational storage and managed pooling
Vector Engine	pgvector	Semantic similarity and distance metrics
Inference	Google Gemini	Primary reasoning and contextual synthesis

---

Quick Start

1. Project Initialization

git clone https://github.com/AyushCoder9/PromptPrep.git
cd PromptPrep/backend
cp .env.example .env

2. Environment Configuration

Populate your .env with the following:

DATABASE_URL="postgresql://postgres.[ref]:[pw]@aws-0-[reg].pooler.supabase.com:6543/postgres?pgbouncer=true"
GEMINI_API_KEY="your_api_key"
GROQ_API_KEY="your_api_key_optional"

3. Execution

# In /backend
npm install && npx prisma generate
npm run dev

# In /frontend (separate terminal)
npm install
npm run dev

---

API Specification

Endpoint	Method	Description
/api/documents/upload	POST	Semantic ingestion and vectorization
/api/quizzes/generate	POST	AI-driven assessment synthesis
/api/flashcards/generate	POST	Conceptual term extraction
/api/qa/ask	POST	Grounded RAG query invocation

---

Built with precision for the modern learner.