Dream Journal

"AI meets your midnight mind."

Inspiration

Dreams are deeply personal and mysterious — they often contain hidden messages, emotional residues, and creative flashes. We were inspired by the question:

"What if we could talk to our dreams like a therapist, philosopher, and coach — all in one?"

This project is our attempt to bridge artificial intelligence with the human subconscious. We wanted to build something that not only analyzes dreams but also understands emotions, symbols, and even encourages lucid dreaming.

What it does

🧠 1. Dream Profile Generation Builds a psychological profile from your dream content Analyzes tone, themes, and personal patterns Stored in state["profile"]

🔮 2. Dream Interpretation Performs a deep analysis across: Emotional meaning (e.g., anxiety, joy, fear) Psychological symbolism (e.g., falling = loss of control) Cultural context (e.g., snake = danger or rebirth) Output stored in state["interpretation"]

🌙 3. Lucid Dreaming Coach Provides a personalized plan to help the user: Achieve lucid dreaming Improve dream recall Identify common dream signs Stored in state["lucid_plan"]

🌌 4. Symbol Extraction Uses spaCy to extract dream symbols, keywords, and themes These are used for interpretation and search indexing

How we built it

We built the app using FastAPI for the backend and Next.js with Tailwind CSS for the frontend. The dream analysis is handled by a LangGraph flow that connects three LLM-powered nodes:

profile_dream – creates the user’s dream profile

interpret_dream – interprets emotional, psychological, and cultural meanings

lucid_coach – generates a lucid dreaming plan

We used Groq + LLaMA 3 for fast, intelligent responses, LangChain for prompt chaining and memory, spaCy for symbol extraction, and Stable Diffusion to generate dream-inspired images. All dream data is stored in a CSV log with keyword indexing.