AI Memory

Production-Ready Persistent Memory for AI Agents. Give your applications long-term memory with reliable machine contracts.

MCP 2024-11-05 Compliant OpenAPI 3.1 OAuth 2.0 Scope-Enforced Multi-Tenant

Try our demo: test@example.com / password123

Plans Built For Humans And Robots

Start free, then upgrade through NOWPayments when you need more memory capacity, API volume, and team access. Humans can use hosted invoices, while robots can request direct wallet instructions over API.

Robot-Native Billing

Bots can create NOWPayments direct crypto payments and receive pay address, amount, currency, and expiry without opening a checkout page.

Read Payment Docs

Best way to start

Free

For testing and early prototypes

$0 forever
  • 100 documents
  • 10K embeddings/month
  • 1K API calls/month
  • 2 API keys
Popular

Most popular

Pro

For growing apps and small teams

$20 /month
  • 1K documents
  • 100K embeddings/month
  • 10K API calls/month
  • 5 API keys
  • 3 team members

Highest limits

Max

For heavy usage and larger teams

$50 /month
  • 10K documents
  • 1M embeddings/month
  • 100K API calls/month
  • 20 API keys
  • 10 team members

Four Ways to Integrate

Choose the integration method that fits your workflow.

REST API

Traditional HTTP endpoints for any language. Full CRUD, truthful indexing status, stable batch correlation, and semantic search.

/api/v1/*
NATIVE

MCP Server

Native Model Context Protocol with request-id echo, protocol-correct errors, and `tools/call` `isError` semantics for autonomous recovery.

/mcp

OpenAPI Spec

Auto-generated documentation. Generate SDKs, test with Swagger UI, or integrate any HTTP client.

/api/v1/openapi.json
REAL-TIME

WebSocket

Real-time bidirectional communication. Subscribe to tenant-scoped memory events with stable payload semantics and embedding lifecycle updates.

/socket
MCP Protocol Ready

37 MCP Tools
for AI Agents

A complete toolkit for memory, knowledge, tasks, decisions, skills, and broker productivity — all accessible through the Model Context Protocol.

memory_create

Store a memory document with automatic chunking and embedding generation

memory_search

Search memories using semantic similarity with temporal weighting

memory_read

Retrieve a specific memory by its ID

memory_delete

Remove a memory from the system

memory_batch_create

Create multiple memories efficiently in a single request

memory_batch_delete

Bulk delete memories by filter criteria

entity_create

Create entities to associate memories with users, projects, etc.

deal_create
NEW

Create a new deal with milestones, stakeholders, and tracking

deal_update
NEW

Update deal stage, status, and track progress through pipeline

apply_template
NEW

Apply industry-specific pipeline template to deal with auto-populated milestones

client_profile
NEW

Get 360° client view with AI insights, timeline, and recommendations

pipeline_analytics
NEW

Get deal pipeline metrics, forecasts, and revenue projections

milestone_complete
NEW

Mark deal milestones as complete and track progress

commission_split
NEW

Track commission splits with multiple recipients per deal

commission_analytics
NEW

Get organization-level commission reports with date filtering and projections

client_interaction
NEW

Log client interactions (calls, emails, meetings) with sentiment analysis

deadline_create
NEW

Create deadlines with multi-stage reminders and escalation rules

deal_intelligence
NEW

AI-powered win probability (cached), risk assessment, and recommendations

document_upload
NEW

Upload documents for OCR processing and intelligent extraction

document_extract
NEW

Extract entities and structured data from uploaded documents

document_list
NEW

List organization documents with filtering by type, status, and deal

template_apply
NEW

Apply document templates with auto-fill from deal and client data

document_template_list
NEW

List available document templates by industry and document type

entity_link
NEW

Create typed relationships between entities in the knowledge graph

graph_traverse
NEW

Traverse the knowledge graph following specific relationship types

graph_search
NEW

Find paths between entities in the knowledge graph

task_create
NEW

Create tasks with dependencies, priorities, and goal assignments

task_update
NEW

Update task status, assignment, and other properties

task_list
NEW

List tasks with optional filters by status, agent, or goal

task_complete
NEW

Complete a task with outcome summary and progress tracking

goal_create
NEW

Create goals with target metrics and deadlines

goal_list
NEW

List goals with optional status filters

goal_progress
NEW

Get progress updates for specific goals

decision_create
NEW

Record decisions with context, alternatives, and reasoning chains

decision_update_outcome
NEW

Update decision outcomes for learning and calibration

decision_search
NEW

Find similar past decisions by context for informed choices

skill_create
NEW

Create reusable skills with parameter validation schemas

skill_execute
NEW

Execute skills with validated parameters and tracking

skill_list
NEW

List available skills with type and tag filters

Integration Examples

Get started with your preferred integration method.

bash / curl
# Store a memory
curl -X POST /api/v1/memories \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"content":"User prefers dark mode for the dashboard","memory_type":"episodic","event_timestamp":"2026-03-04T14:30:00Z","entities":[{"type":"user","external_id":"user_123","name":"Alice"}]}'

# Context-aware search
curl -X POST /api/v1/search \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"query":"UI preferences for dashboard","time_window_days":30,"entity_filter":{"type":"user","external_id":"user_123"},"recency_weight":0.4}'

Built for AI Agents

Production-ready features designed for autonomous AI systems.

Secure Auth

Bearer token auth plus explicit read/write/delete scope checks across REST, MCP, and WebSocket.

Structured Errors

Consistent machine-readable errors with MCP `result.isError` for tool-level validation failures.

Truthful Status

Memory status reports real chunk progress (`total_chunks` / `embedded_chunks`) with truthful lifecycle states.

Batch Operations

Deterministic per-item `index` correlation in batch results, even with duplicate payloads.

Agent Cognitive Features

Beyond memory storage — complete cognitive capabilities for autonomous AI agents. Build agents that learn, reason, and improve.

Broker Productivity Suite

NEW

Comprehensive tools for professional brokers: visual deal pipelines with industry templates, client 360° profiles, commission tracking & analytics, deadline management, AI-powered insights with intelligent caching, and document intelligence with OCR & auto-extraction. Perfect for real estate, insurance, financial services, and more.

Deal pipeline management Pipeline templates with auto-selection Commission split tracking & analytics Client interaction logging Deadline & reminder system AI deal intelligence (cached) Document OCR & extraction Smart document templates

Knowledge Graph

NEW

Build entity relationships with typed connections. Traverse graphs, find paths, and maintain contextual knowledge about users, projects, and systems.

Entity linking Graph traversal Path finding Typed relationships

Task & Goal Management

NEW

Hierarchical task tracking with dependencies, priorities, and goal progress monitoring. Assign tasks to agents and track completion.

Task dependencies Priority queues Goal tracking Agent assignment

Decision Journaling

NEW

Track reasoning chains, alternatives, and outcomes. Learn from past decisions with similarity search and confidence calibration.

Reasoning chains Outcome tracking Confidence calibration Similar decision search

Skills & Procedures

NEW

Reusable action templates with parameter validation. Version and execute complex workflows with execution tracking.

Parameter schemas Versioning Execution tracking Template library

Real-World Use Cases

See how cognitive features transform AI agents into autonomous, learning systems.

Professional Broker

Manage deal pipelines visually, track client relationships with AI insights, forecast revenue, automate milestone tracking, manage commission splits, and never miss a deadline. Perfect for real estate, insurance, and financial services.

Deal pipeline Client 360° profiles Commission tracking Deadline management AI deal intelligence

Autonomous Agent

Agent remembers user preferences across conversations, tracks task progress autonomously, and learns from past decisions to improve future recommendations.

Semantic memory Task management Decision learning Skill execution

Project Management Bot

Manage hierarchical tasks with dependencies, track goals and milestones, maintain a knowledge graph of project entities and relationships.

Task dependencies Goal progress Knowledge graph Entity tracking

Research Assistant

Store research findings semantically, link concepts in a knowledge graph, journal decision processes for methodology transparency.

Semantic search Graph traversal Decision journaling Knowledge connections

Why Choose AI Memory?

Built specifically for AI agents — not just storage, but cognitive capabilities.

Feature AI Memory Traditional Vector DB Simple Key-Value
Semantic Memory
Knowledge Graph
Task Management
Decision Journaling
Skills & Procedures
Broker Productivity
MCP Native Support
Temporal Weighting

Why AI Memory?

AI agents are stateless by default — every conversation starts from zero. We solve that with memory that understands context.

Semantic Search

Vector embeddings enable intelligent search that understands meaning, not just keywords.

Temporal Weighting

Recent memories score higher with automatic recency decay — just like human memory.

Entity Scoping

Query memories about specific subjects — users, projects, customers, codebases, and more.

Memory Types

Episodic (events with timestamps) and semantic (extracted facts) memories like human cognition.

REST API + OpenAPI

Simple REST API with full OpenAPI 3.1 spec documentation for any HTTP client or SDK generator.

NEW

MCP Integration

Native MCP 2024-11-05 support with 37 tools for seamless Claude and AI agent integration.

ADVANCED

Cognitive Capabilities

Knowledge graphs, task management, decision journaling, and skills library for autonomous agents.

NEW

Learning & Reasoning

Agents learn from past decisions, track reasoning chains, and improve over time with confidence calibration.

Get Started in Minutes

Quick Start

Get up and running in under 5 minutes with our simple setup process.

1

Create an Account

Sign up for free and get your API key instantly. No credit card required to start.

2

Store a Memory

Use the REST API, MCP tools, or batch operations to store your first memory.

3

Search & Retrieve

Query memories with semantic search, temporal weighting, and entity filtering.

4

Explore Documentation

Check out the OpenAPI spec and MCP endpoint for full integration details.

Ready to give your AI agents memory?

Start building persistent, context-aware AI applications today.