Memlayer

The plug-and-play memory layer for smart, contextual agents

Memlayer adds persistent, intelligent memory to any LLM, enabling agents that recall context across conversations, extract structured knowledge, and surface relevant information when it matters.

<100ms Fast Search • Noise-Aware Memory Gate • Multi-Tier Retrieval Modes • 100% Local • Zero Config

Links

• GitHub: github.com/divagr18/memlayer
• PyPI: pypi.org/project/memlayer

Quick start

Install:

pip install memlayer

Basic usage:

from memlayer.wrappers.openai import OpenAI

client = OpenAI(model="gpt-4.1-mini", storage_path="./memories", user_id="user_123")
client.chat([{"role": "user", "content": "My name is Alice and I work at TechCorp"}])
response = client.chat([{"role": "user", "content": "Where do I work?"}])
# -> "You work at TechCorp."

Memlayer automatically filters, extracts, stores and retrieves relevant memories, no manual prompts required.

---

What makes Memlayer different

• Selective long-term memory — we only store what matters, not every chat line.
• Hybrid storage — semantic vectors for recall plus a lightweight knowledge graph for relationships and updates.
• Noise-aware gate — cheap salience checks (prototype embeddings + TF-IDF) keep the store clean.
• Multi-tier retrieval — Fast / Balanced / Deep modes so you can trade latency for depth.
• Offline-first — runs locally with no external services required.
• Zero config — drop it into your app and start remembering.

---

Explore the docs

• Basics & Quickstart — architecture and getting started
• Search Tiers & Modes — how retrieval and salience modes work
• API Reference — full method docs and examples
• Providers — OpenAI, Claude, Gemini, Ollama, and local model tips
• Storage Backends — ChromaDB + NetworkX details

Or open the sidebar to browse everything.

---

Examples & development

See examples/ for runnable demos (basics, search tiers, provider samples).
Clone and run locally:

git clone https://github.com/divagr18/memlayer.git
cd memlayer
pip install -e .
python examples/01_basics/getting_started.py

---

Contributing

Bug reports, PRs, and documentation fixes are welcome. See the repo CONTRIBUTING.md for guidelines.

---

If you want a slightly shorter or more technical landing page (or want the badges moved here), tell me which tone you prefer and I’ll adapt it.