Memory,
by Design.

Percolation Labs builds the intelligent memory layer that gives AI systems context, continuity, and recall — so every interaction feels like it actually remembers.

Explore Our Work See Products
Our Mission

The memory layer
AI has always needed

Today's AI is brilliant but it takes too much work to integrate your content and your rules. Percolation Labs engineers the persistent, structured memory infrastructure that transforms stateless models into agents that truly know you.

Our Capabilities
Recall Context Pattern Continuity Store Index AI Memory
Core Technology

Built for memory,
from the ground up

Every component of our stack is designed specifically for the constraints and demands of AI memory — not adapted from general-purpose tools.

01
Storage Layer

Semantic
Storage

A purpose-built vector, entity, and graph store that preserves meaning, not just text. Moments are stored with semantic richness, temporal metadata, and relational context.

02
Reasoning Layer

Temporal
Reasoning

Memory is inherently temporal. Our system understands how your concerns and your projects are changing over time — not just timestamps.

03
Security Layer

Privacy
Architecture

Encrypted at rest and in transit. User-controlled deletion. No model training on personal data. Privacy isn't a policy — it's a constraint baked into the architecture.

Our Products

Two ways to experience
Percolation memory

RemApp home feed
RemApp chat
Consumer App

RemApp

The personal memory experience. Capture moments like a feed, chat with your past, and let AI surface the patterns you didn't know were there.

Moment Feed

Add anything worth remembering — text, images, links — indexed instantly.

Chat with your past

Ask questions. Get insights. Your history as a thinking partner.

Pattern Discovery

AI finds connections across time — themes, habits, turning points.

Download on the App Store Get it on Google Play
Enterprise

P8 Memory
Cloud

The enterprise memory infrastructure. Deploy Percolation memory on our cloud or your own — with full control over data sovereignty, agents, and ontology.

Your cloud or ours

Full deployment flexibility — SaaS or private cloud, your infrastructure, your rules.

Custom agents & ontology

Bring your own agents and domain knowledge, or build on ours. The memory layer adapts to you.

Data sovereignty & open source

Fully open source with K8s deployment recipes. Air-gapped deployments, enterprise SSO, and compliance controls.

Talk to Us Open Source · MIT
How It Works

Three steps to a
living memory

1

Capture

Add any content — documents, conversations, notes, data streams — to your Percolation memory. Run on our cloud or deploy entirely on yours. Anything worth remembering belongs here.

2

Understand

Our agents build context, extract structure, and evolve a knowledge graph around your memory. Use ours out of the box, or bring your own agents, ontologies, and domain logic to make it truly yours.

3

Surface

Put your memory to work — directly in Claude Desktop, ChatGPT, or any chat application via our MCP and API integrations. Or experience it through RemApp, our purpose-built memory interface.

Work with us

Memory deserves
to be designed.

Whether you're building an AI product that needs persistent context, or you want to experience RemApp — we'd love to talk.

Get In Touch Explore RemApp