Skip to main content

Getting Started

A practical guide to using Vidurai in your projects.

Basic Setup​

from vidurai import Vidurai

# Initialize with defaults
memory = Vidurai()

# Store a memory
memory.store("User prefers dark mode interface")

# Recall memories
results = memory.recall("interface preferences")
print(results)

Understanding Memory Flow​

  1. Store: Add new information
  2. Compress: Automatic when tokens exceed limits
  3. Recall: Retrieve relevant information
  4. Decay: Gradual importance reduction over time

Common Patterns​

Pattern 1: User Preferences​

# Store user preferences
memory.store("User works as a Python developer", importance=0.9)
memory.store("User prefers type hints in code", importance=0.8)

# Recall when generating code
prefs = memory.recall("coding preferences")

Pattern 2: Conversation Context​

# Store conversation turns
for msg in conversation:
    memory.store(f"{msg['role']}: {msg['content']}")

# Get relevant context
context = memory.recall(query="project requirements", limit=5)

Pattern 3: Knowledge Base​

# Store permanent knowledge in Vijnanamaya Kosha
memory.store(
    "Company coding standards: Use black formatter, pytest for tests",
    importance=1.0  # Never forgotten
)

Next Steps​