Governed memory infrastructure
Memory your agents can trust.
Store durable knowledge, enforce isolation, and review every fact before your agents recall it.
$ memhouse query --account acme --scope support
fact customer prefers annual billing
state active
source ticket_4817
account acme
A control plane for what agents remember.
Vector search retrieves similar text. MemHouse governs durable facts through identity, provenance, lifecycle state, and policy.
Hard account boundaries
Bind every query to verified account and scope claims.
Complete provenance
Trace recalled facts to their source, model, and extraction time.
Review before recall
Hold sensitive or uncertain facts for explicit human approval.
Elixir engine, open protocols
The Elixir high-concurrency engine runs entirely within your own cloud infrastructure (VPC), with REST today and Python and packaged MCP clients coming soon.
How MemHouse Governs Agent Memories
Unlike raw vector databases that blindly ingest embeddings, MemHouse filters, isolates, and subjects extracted facts to explicit governance lifecycle states.
Raw Observation Input: Multimodal Ingest
Unstructured prompt logs, customer support chats, document snippets, or API webhooks are ingested with tenant Account ID and Scope headers.
Verified exact account boundary match with high confidence.
Fact matched plan registry schema.
Flagged for Gate A Human-in-the-Loop review: PII preference update requires supervisor confirmation.
Built for Personal Agents, Frameworks & Swarms
Whether powering standalone personal AI companions or serving as the external persistent state memory for multi-agent swarms, MemHouse provides governance, speed, and hard isolation.
OpenClaw Integration
Coming soon: packaged OpenClaw and desktop-agent connectors. Use the documented HTTP API today; the server's MCP endpoint is experimental.
Hermes Agent Memory
Coming soon: a Hermes Agent adapter for governed long-term preferences and codebase knowledge. Use the HTTP API for current experiments.
Privacy-Scoped Memory
Keep sensitive personal facts (API keys, private keys, health info) cryptographically bound to your user account scope. Zero risk of cross-account data leakage.
Stateless agents, stateful engine.
Keep your agent prompts lightweight. Pass observation payloads to MemHouse, and query governed memories through the HTTP API. SDK and packaged MCP examples are clearly marked when they are illustrative or coming soon.
# Python SDK coming soon. Use the shipped HTTP API today.
import requests
base_url = "https://memhouse.internal.company.com"
headers = {"Authorization": "Bearer YOUR_API_KEY"}
response = requests.post(
f"{base_url}/api/v1/context",
headers=headers,
json={"scope_path": "/support/tickets/INV-9042", "budget_chars": 4000},
)
response.raise_for_status()
print(response.json()["data"])
Why Enterprise AI Teams Choose MemHouse
Vector databases index text; MemHouse governs knowledge. Compare our stateful memory engine against dedicated memory servers and raw vector DBs.
| Capability / Requirement | MemHouse | Zep | Mem0 | Letta (MemGPT) | Postgres + pgvector | Pinecone |
|---|---|---|---|---|---|---|
Human-in-the-Loop Curation (Held State) Category: Governance | Built-in (Gate A & B) | None | None | Manual code rules | Must build from scratch | None |
Strict Account & Scope Boundaries Category: Security & Isolation | Identity-bound account and scope checks | Session metadata filters | Basic user_id metadata | Single tenant core | Requires custom RLS SQL | Namespace metadata |
Fact Lifecycle States (Held / Provisional / Active) Category: Governance | 4-Stage State Machine | Flat vector list | Raw memory blocks | Archival vs Core blocks | None | None |
Verifiable Audit Provenance & Citation Metadata Category: Governance | Complete Audit Trail | Basic timestamps | Basic timestamps | Memory log history | Custom table schema required | Vector metadata payload |
Runtime Backend Architecture Category: Architecture | PostgreSQL pgvector & Async Engine | Go / Python | Python | Python | N/A | Proprietary Cloud |
Zero-Dependency VPC Air-Gapped Deployment Category: Security & Isolation | Coming soon | |||||
Model-Context-Protocol (MCP) Native Client Category: Developer Experience | Experimental endpoint; packaged client coming soon |
Last verified August 16, 2026 against public product documentation from Zep, Mem0, Letta, PostgreSQL, and Pinecone. Competitor capabilities and packaging change; verify current vendor documentation before making a purchasing decision. “Coming soon” identifies MemHouse capabilities that are not part of the current released product.
Deploy on your terms.
Start with the source-available fair-code engine, operate inside your VPC, or join the managed cloud waitlist.
Source-Available Fair-Code
Community Edition
$0
forever free
Independent Developers, AI Startups & Fair-Code Adopters
- Core memory engine
- PostgreSQL + pgvector database integration
- Basic Gate A quality filter & Gate B account scoping
- REST API; Python SDK and packaged MCP client coming soon
- Docker Compose single-command self-hosting
Governed VPC Infrastructure
Enterprise Edition
Custom
billed annually
Regulated Enterprises, CISOs, Multi-Tenant SaaS & Financial/Legal AI
- Everything in Community, plus:
- Full VPC & On-Premise deployment (Zero phone-home)
- Advanced Human-in-the-Loop Curation Web Console
- Identity-bound Account & Scope authorization controls
- Custom Gate A governance policy rules engine
Managed Cloud Service
Cloud Waitlist
Usage-Based
pay per 10k ops
SaaS Teams & AI Apps seeking Zero-Ops Managed Memory
- Fully managed pgvector serverless cluster
- Instant API Keys & regional edge routing
- Automated backup, replication & zero downtime updates
- Built-in Gate A & Gate B web interface
- Usage-based pricing with generous free monthly quota
Make agent memory accountable.
See how MemHouse enforces isolation, provenance, and human approval for security-sensitive deployments.