GibsonAI
GibsonAI, operating as Memori Labs Inc. (DBA Memori), builds LLM-agnostic memory infrastructure that captures, classifies, and retrieves structured context for production AI agents, reducing LLM token costs by 95%. It serves AI agent developers and enterprise AI deployers via a freemium SDK, Memori Cloud, and direct sales.
- Company typePrivate
- Founded2024
- HeadquartersNew York, United States
- Headcount1–10
- GTM typeB2B
- OfferingSoftware
What GibsonAI does
GibsonAI, operating under the DBA "Memori" as the legal entity Memori Labs Inc., is a New York-based private company founded in 2024 that builds agent-native memory infrastructure for production AI systems. The core product, Memori, is a LLM-agnostic layer that captures each agent chat turn and automatically classifies it into facts, preferences, rules, and summaries, then performs targeted recall and selective semantic search to inject only relevant context on demand. Supporting modules include Memory Graph (entity relationship visualization), Memori Analytics (usage and cache hit rate monitoring), Memori Cloud (hosted deployment launched 2025), and Payments Vault (PCI/SOC 2-compliant handling of payment data and PII). Published benchmark results report 81.95% accuracy on the LoCoMo benchmark with a 95% reduction in token usage versus full-context retrieval.
The company serves two primary segments: AI agent developers and builders integrating memory into production systems, and enterprise AI deployers requiring secure, scalable memory infrastructure with cost controls. A secondary segment targets robotics and IoT use cases. The go-to-market combines product-led growth (one-line SDK integration, free tier, sub-minute onboarding) with community-led motion (open-source GitHub SDK, Discord, blog, benchmark research) and direct enterprise sales for larger deployments. Revenue is generated through a freemium model with subscription-based Pro tier, usage-based scaling on paid plans, and a startup program offering 3-month free Pro access. Notable disclosed partnerships and endorsements include MongoDB, Cockroach Labs, DigitalOcean, SK Networks (NAMUHx wellness robots), Swarms Corporation, and Agno.
The company is in an early commercial stage with 1-10 employees and has raised a single disclosed funding round — a $3.5M seed in May 2024 led by Oceans with participation from f7 Ventures, RiverPark Ventures, and Struck Capital. No revenue, ARR, or paying customer count is publicly disclosed. Compliance posture includes PCI DSS and SOC 2 Type 2 certifications, positioning the company for regulated enterprise workloads ahead of typical early-stage timelines.
GibsonAI firmographics
Firmographics- Name
- GibsonAI
- Legal name
- Memori Labs Inc.
- Website
- https://gibsonai.com
- Company type
- Private
- Founded year
- 2024
- Operating status
- Operating
- Headcount range
- 1–10 employees
- Short description
- GibsonAI, operating as Memori Labs Inc. (DBA Memori), builds LLM-agnostic memory infrastructure that captures, classifies, and retrieves structured context for production AI agents, reducing LLM token costs by 95%. It serves AI agent developers and enterprise AI deployers via a freemium SDK, Memori Cloud, and direct sales.
- Ownership category
- akta.pro rank
GibsonAI industry classification
Industry- Product category
- AI Memory Infrastructure
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518)
- SIC
- Services-Prepackaged Software (7372)
- akta.pro primary industry
- LLMOps & Generative AI Platforms (Prompt/Agent Orchestration, RAG) (HDAEANAD)
- akta.pro secondary industries
- AI Memory & Storage for Training (HBM/DDR/SSD/Object Storage) (HDAAAAAF), Database-as-a-Service (DBaaS) Platforms (HDAEAAAN)
Keywords
Where GibsonAI is headquartered
LocationHeadquarters
- HQ city
- New York
- HQ country
- United States
- HQ region
- North America
Markets served
GibsonAI business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations
Revenue model
- Memori Cloud Subscription: Cloud-hosted memory storage and search service with tiered plans including Pro tier for startups (3-month free access program available). Usage-based scaling for production workloads.
- Self-serve SDK: Developer SDK with free signup tier, enabling tokenless recall and structured memory. Revenue generated through usage-based consumption on paid tiers.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Freemium | Monthly | Free tier with basic memory infrastructure |
| Subscription | Monthly | Pro plan for growing teams |
Go-to-market motion2 records
Distribution channels4 records
Marketing channels8 records
GibsonAI product offering
Product offeringCore offering
GibsonAI (operating as Memori Labs Inc., DBA Memori) provides an LLM-agnostic memory infrastructure layer that turns AI agent execution and conversation into structured, persistent state for production systems. Its product captures each chat turn, classifies content into facts, preferences, rules, and summaries, and serves targeted recall with semantic search to reduce LLM token costs. The offering includes a Python SDK for drop-in integration and a hosted Memori Cloud service.
Product overview
Memori Labs offers a unified platform for agent-native memory infrastructure. The core product is Memori, a LLM-agnostic memory layer that transforms agent conversations into structured, persistent state. The platform includes Memory Graph for visualizing entity relationships, Memori Analytics for monitoring performance metrics, and Payments Vault for secure handling of payment data and PII. Memori Cloud provides a hosted deployment option with managed infrastructure, semantic search, and graph relationship capabilities. Organizations can self-host by storing memory in their own database while layering on Memori's hosted capabilities.
Differentiator
Problem solved
Functional benefit
Products and services
- Memori LLM-agnostic memory infrastructure SDK that captures AI agent chat turns, classifies them into facts, preferences, rules, and summaries, and provides targeted recall and semantic search across conversations and documents. Designed for developers building AI agents and enterprises deploying production AI systems requiring persistent memory.
- Memori Cloud Hosted cloud deployment of the Memori memory layer providing instant memory storage and search with managed infrastructure, semantic search, memory graph relationships, analytics, and third-party integrations. Targets teams that prefer managed infrastructure over self-hosting the SDK.
Quantifiable outcome
- 81.95% accuracy on LoCoMo benchmark while reducing token usage by 95%
- +2 more outcomes
Companies that use GibsonAI
Customer profileNamed customers6 records
Segments3 records
Ideal customer profiles3 records
GibsonAI technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration2 records
AI capability4 records
Feature9 records
GibsonAI partnerships and signals
Strategic signalPartnerships
Three partnerships are on record, tiered core.
- MongoDBcoreMongoDB is exploring integration of Memori's memory abstraction with MongoDB's flexible document model and integrated vector search. Andrew Davidson, SVP Products, cited the combination as a powerful approach for richer, more dynamic context in LLM-driven applications.
- Cockroach LabscoreCockroachDB integration enables developers to spin up memory-backed AI workflows with minimal cost and friction. Spencer Kimball, Co-founder & CEO, endorsed the integration for testing and production scaling without upfront database costs.
- SK NetworkscoreSK Networks uses Memori for NAMUHx wellness robots, providing persistent memory layer for truly personalized wellness experiences. Simon Hong, Head of AI Innovation, cited Memori's strong recall accuracy and inference-cost savings as key factors.
Scale indicators3 records
Recent moves6 records
Expansion highlights7 records
GibsonAI competitors and assessment
Company assessmentDirect peers
- Mem0: Mem0 is a direct competitor offering a memory layer for LLM applications and AI agents, with automatic memory extraction, storage and retrieval across sessions. It targets the same developer persona, uses a similar self-serve/SDK model and competes head-to-head on accuracy benchmarks and token-cost reduction.
- Zep (GetZep): Zep provides long-term memory and a temporal knowledge graph for AI assistants and agents, with APIs for message storage, summarization and fact extraction. It overlaps closely with Memori on agent memory infrastructure and on the use of structured recall rather than raw context stuffing.
- Letta: Letta (formerly MemGPT) builds persistent memory and stateful agent infrastructure for LLM applications. It competes directly on long-horizon agent memory, context management and developer-facing APIs for production deployments.
- Cognee: Cognee offers an open-source memory layer for AI applications that structures and retrieves knowledge from conversations and documents using graph and vector storage. It is comparable to Memori on the agent-memory abstraction, structured recall and self-hostable architecture.
Broad incumbents
- LangChain: LangChain is the dominant LLM application framework with its own agent orchestration, memory abstractions and integrations across hundreds of vector stores. It is a broad incumbent whose native memory components overlap with Memori's value proposition for the same developer audience.
- LlamaIndex: LlamaIndex provides data frameworks for LLM applications including indexing, retrieval and agent memory components. It is a broad incumbent in the same ecosystem, offering overlapping capabilities around structured context and agent-side recall.
- Pinecone: Pinecone is a leading managed vector database used for retrieval-augmented generation and agent memory use cases. It is a broad incumbent with significantly more capital and a larger installed base that could expand into the agent-memory category Memori occupies.
- Weaviate: Weaviate is an open-source vector database with hybrid search, modules and a managed cloud service widely used for RAG and agent memory. It is comparable on the retrieval and semantic-search backbone that Memori layers its agent-memory product on top of.
Emerging players
- Chroma: Chroma is an open-source embedding database popular with AI developers for RAG and lightweight agent memory. It overlaps with Memori on the developer-targeted self-serve distribution and on serving as a retrieval layer underneath agent applications.
- Qdrant: Qdrant is an open-source vector similarity search engine with a managed cloud offering, used for semantic search and memory in LLM applications. It is comparable as a retrieval infrastructure layer that Memori's customers and competitors rely on for agent context.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks6 records
Key highlights6 records
Customer concentration
GibsonAI social profiles
Digital presenceGibsonAI compliance and trust
Trust signalCompliance2 records
GibsonAI financial estimates
Financial estimateRevenue estimate
Valuation estimate
GibsonAI leadership team
Management profileNumber of profiles
GibsonAI funding detail
Funding detailFunding overview
Funding rounds1 record
Investors4 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
GibsonAI M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about GibsonAI
What does GibsonAI do?
GibsonAI (operating as Memori Labs Inc., DBA Memori) provides an LLM-agnostic memory infrastructure layer that turns AI agent execution and conversation into structured, persistent state for production systems. Its product captures each chat turn, classifies content into facts, preferences, rules, and summaries, and serves targeted recall with semantic search to reduce LLM token costs. The offering includes a Python SDK for drop-in integration and a hosted Memori Cloud service.
Is GibsonAI a public or private company?
GibsonAI is a private company. It is classified as venture growth investor backed and is currently operating.
When was GibsonAI founded?
GibsonAI was founded in 2024. It employs 1 to 10 people.
Where is GibsonAI based?
GibsonAI is headquartered in New York, United States, in the North America region.
How does GibsonAI make money?
Two revenue lines are on record. Memori Cloud Subscription is the primary driver. The others are self-serve SDK.
Who are GibsonAI's main competitors?
Direct peers on record are Mem0, Zep (GetZep), Letta and Cognee. Broad incumbents are LangChain, LlamaIndex, Pinecone and Weaviate. Emerging players are Chroma and Qdrant.
Does GibsonAI have an API?
Yes. Drop the SDK into existing code to handle model calls and callbacks with zero configuration. The SDK enables developers to integrate Memori's memory infrastructure into their AI agent workflows. Developer documentation is at memorilabs.ai/docs.
What industry is GibsonAI in?
GibsonAI's product category is AI Memory Infrastructure. Its primary akta.pro industry code is HDAEANAD, LLMOps & Generative AI Platforms (Prompt/Agent Orchestration, RAG), with a secondary code of HDAAAAAF, AI Memory & Storage for Training (HBM/DDR/SSD/Object Storage). Its NAICS code is 518 and its SIC code is 7372.