SOC 2 compliant agent memory systems

Universal memory layer for AI agents. Three concurrent stores (vector + graph + KV); LLM-extracted facts; concurrent retrieval via ThreadPoolExecutor.

Acquired by ClickHouse (2026-01-16); product and MIT-licensed repo continue under the Langfuse name. Memory module reads/writes captured as named spans. Trace Log View concatenates every agent step including memory ops. Agent Graphs (GA …

Memory reads, vector DB retrievals, state changes are distinct span types in traces. RAG eval separates retrieval quality (context precision) from generation quality (faithfulness). Dataset versioning guards against eval drift.

Bi-temporal knowledge graph (event time + ingestion time). Strong on chronological reasoning and contradiction tracking. Graphiti is the open-source core.

Cloud browser infrastructure for AI agents. Headless Chromium sessions with persistent cookies, localStorage, session recording — agent workflows carry state across runs. Memory at the session/automation tier rather than conversational.

Standalone memory service from Vectorize. Open source. Biomimetic four-network design (World, Bank, Observation, Opinion). Ships an MCP memory server.

Structured Knowledge items stored server-side. Three-tier scope: repo-pinned (always injected), global (recall by relevance), org/enterprise (visible to org members). Auto-generates Repo Knowledge from repo structure each session; reads …

Formerly an NPC character engine with named Long-Term Memory; now sells realtime voice models (TTS-2, STT-1, speech-to-speech) and OpenAI-compatible open-model inference. Character/NPC memory is no longer the headline product.

Acquired by Check Point Software (completed by 2025-11-11); being folded into Check Point's AI security platform. Screens every prompt, response, and retrieved document for indirect prompt injection — primary vector for memory poisoning.…

Memory engine with API, app, browser extension, and MCP server. Extracts facts, tracks updates, resolves contradictions, auto-forgets expired info. Plugins for Claude Code, OpenCode, OpenClaw, Hermes.

Acquired by Proofpoint (closed 2026-02-12). Runtime enforcement targeting memory poisoning, unauthorised execution, identity spoofing per the OWASP LLM threat list. Visibility/control over MCP servers and locally-installed AI tools — the…

Real-time knowledge graph indexed from connected repos + commit history + CI/CD + tickets + docs. Separate Agent Memory layer (Sept 2025) captures session decisions/corrections/goals as discrete items with developer review/approval UI. M…

Mac/iOS app (no hardware). Captures meeting audio via system-audio tap (no bot joining); generates notes from your shorthand + audio. Local audio processing claim; cloud for summaries. Enterprise tier opts out of model training.

AI security platform with four modules: AI asset discovery, supply-chain/model integrity validation, continuous attack simulation, and runtime detection/response, including protection of agents and MCP-based systems.

No documented built-in persistent semantic memory store. Session duration expanded to 200 min (Agent 3); Agent 4 shifts toward persistent runtimes. Temporal used for orchestration durability, not semantic memory. Cross-session memory rel…

Clinician-assist ambient documentation. Source mapping: every AI-generated summary element traced back to the source utterance. Audit-and-trust layer over episodic memory. Built on proprietary 1.5M+ medical-encounter dataset.

Memory tool API + consumer "Memory" feature + Auto Dream consolidation. Document-as-memory rather than extracted-fact-as-memory.

Treats memory as first-class architecture. Captures the digital footprint of every interaction; retrieves preference and history at engagement time. Public example: airline knowing a frequent flyer wants aisle seats with her son — prefer…

Per-plan vector-DB storage quota + LLMz autonomous engine (in-session working memory) + Knowledge Base (semantic search over uploaded docs). Long-term user memory persists across sessions.

Agentic AI platform for life sciences: orchestrated scientific agents over a curated biomedical knowledge graph (plus Bio Graph and Pipeline Graph), an evidence-retrieval system and an enterprise data fabric over literature, trials and p…

Acquired by NiCE (closed 2025-09-08); continues as NiCE Cognigy, standalone and inside CXone Mpower. Enterprise contact centre (voice + chat). Short-term covers current session; long-term persists across sessions and is shared by every a…

Mimir hierarchical memory: short-term (verbatim), medium-term (summaries + emotion), long-term (cross-session). Scene awareness layer adds in-engine perception. Unity + Unreal SDKs with real-time voice + lipsync.

Hierarchical block-based documents + nested pages (no native graph). In-document AI assistant: summarise, rewrite, translate, Q&A over docs, proofread.

Persistent instructions injected at start of model context: Project Rules (.mdc files in .cursor/rules, scoped Always/Auto Attached/Agent Requested/Manual), global User Rules, and AGENTS.md; legacy .cursorrules deprecated. Codebase index…

Autonomous support agents resolving end-to-end (refunds, cancellations). Channel-agnostic memory: chat → call carries prior context. Agents recognise returning customers and recall prior issues.

Specialty-care ambient scribe (oncology focus). Pulls forward prior notes with precise dates; synthesises interval history (what changed since last visit); auto-populates disease-status and treatment-plan context.

The Browser Company's AI-native browser; Atlassian acquired The Browser Company for $610M cash (announced 2025-09, closed 2025-10-21). Stores summaries and preferences locally; opt-in @history feature references up to 7 days of browsing …

Embedded in case-management platform. Holds full scope of every matter — deadlines, billing, discovery, communications, negotiation history — as live context for every query. Case data is the context, not a separate retrieval corpus.

Acquired by Cisco (announced 2026-04-09, closed by 2026-05); now Splunk Agent Observability. Treats memory as first-class in multi-agent tracing. Luna-2 SLMs (3B / 8B) scan every interaction for intent drift and belief drift; 20+ checks …

Gemini app learns from past chats via the Personal context setting (on by default, can be disabled; temporary chats excluded). Users can import memories and full chat history from ChatGPT, Claude and other assistants (18+, personal accou…

Memory carries forward user-chosen context (matter details, precedent, preferences, approved practices) and reuses past threads scoped to a user, client, matter or firm process. Firm admins enable/disable it; ethical walls and audit cont…

Patient-conversation / chronic-care staffing model. HIPAA-compliant Memory Store holds non-EHR salient facts from prior calls, persists across engagements, explicitly not used for model training. Shifts from transactional call-handling t…

Memory is a file explorer for what Lindy knows: files it creates or uses, files users upload, skills (.skills/ folders with SKILL.md) and durable knowledge in a .memory/ area. Scoped Personal (owner only), Team (admins edit, members read…

Production memory infrastructure powering LinkedIn's Hiring Assistant and other GenAI apps. Three-layer (episodic/semantic/procedural) shared memory substrate across multi-agent systems; recent-context retrieval + semantic search + summa…

Jan 2026 platform relaunch. Persists negotiation history, prior decisions, accepted risk positions across all contracts; new terms align with what the org has already agreed. Branded as the antidote to "enterprise amnesia."

Teacher productivity platform (80+ teacher tools, 50+ student tools) for planning, differentiation, IEP generation, feedback. Does not maintain a cross-session learner model. Teachers input student context manually each session; district…

Commercial Mem0 ships SOC 2 / HIPAA, zero-trust access controls, BYOK encryption, real-time monitoring, audit logs and workspace governance as defaults (Trust Center at trust.mem0.ai). For privacy-first deployments the self-hosted path i…

Microsoft 365 Copilot memory (saved memories, chat-history details, custom instructions) stored in a hidden folder of the user's Exchange mailbox under mailbox compliance; on by default, admin-controllable; available to Copilot Chat user…

Copilot Mode retired 2026-05-13; its features now ship as Copilot in Edge. Journeys clusters browsing history into topic cards on the new tab page (desktop, and US mobile); Copilot long-term memory can reference past chats and browsing h…

Mistral's Agents API provides persistent conversation state across turns and sessions; runtime maintains branching conversation trees. Memory is implicit in the stateful session model rather than a separate store — agents can branch, con…

Le Chat is now Vibe; Mistral's help center describes Memories as a Vibe feature (article dated 2026-08-12). Memories in Mistral's consumer assistant: saves details on user request or automatically; user can view/edit/delete or clear all;…

Vibe is Mistral's cloud-hosted coding-agent platform: each session runs in an isolated sandbox with task state, history, and approvals. Spawnable from CLI or Le Chat, with parallel execution. Local CLI sessions can be "teleported" to the…

Low-code workflow platform. AI Agent node ships with pluggable memory sub-graph: Simple Memory for development; Postgres/Redis for production; vector store nodes (Qdrant, Pinecone, MongoDB Atlas) for semantic recall.

Ambient scribe + longitudinal clinical intelligence (strategic partnership Jul 2025). Live encounter transcript reconciled against historical labs/imaging/notes; condition-suspecting + care-gap engine runs against the reconciled record.

Notion's first-party AI. Reads up to 50 pages per query (as of Jan 2026); Custom Instructions persist brand voice / team structure / preferences. Context follows you across model swaps. Notably not persistent cross-session learning — it …

Wearable orb (necklace or clip) with open-source hardware + firmware. Continuous audio capture; real-time GPT-4o transcription; searchable memory graph. Supports app screen capture too. Hybrid privacy: cloud by default (SOC 2/HIPAA/AES-2…

First-party ChatGPT memory combining saved memories, reference to past chats, custom instructions and connected apps; 'improved memory' keeps an evolving summary (legacy list-style saved memories still selectable), with review/correct/de…

Memory under ~/.codex/memories/ : summaries, durable entries, recent-input logs, supporting evidence — generated in background after sessions go idle. Chronicle (opt-in macOS preview) uses screen-capture to seed memory. Off by default; o…

Joins Zoom / Meet / Teams as a meeting bot. Real-time transcription, speaker identification, searchable archive across past meetings. SSO + admin controls at Business tier.

AI-native Chromium-based browser. Assistant operates across tabs and sessions; persistent memory of preferences; multi-step agentic tasks without leaving the browser. Memory is opt-in and server-side; enterprise memory tier separately av…

Stores preferences and details (and optionally search history) to personalize answers, citing which memory was used; user controls in Settings → Memory. 'Brain' adds structured self-improving memory for Max subscribers and shared project…

Magnetic clip-on wearable. Records in-person conversations and calls; AI summaries, action items, searchable transcripts. AES-256 + TLS 1.2 in transit; doesn't retain audio after transcription. Also sells Note (card-form recorder) and No…

Meeting bot + browser extension. Captures audio + video; engagement scoring (talk-time, sentiment, attention); enterprise search across all past meetings. "Ada" digital-twin feature builds an AI memory of work patterns.

Two memory scopes: conversation metadata (KV per-conversation) and persistent memory (cross-session). Enabling metadata auto-adds Add/Read conversation metadata tools.

Bret Taylor + Clay Bavor. Agent Data Platform links unstructured conversation data to enterprise systems (billing, inventory, policies, transactions); persists across channels and time. Memory accumulates across every interaction, not ju…

Doctor-assist ambient documentation. Pulls forward prior documentation into a pre-visit synthesis rather than waiting for the encounter. Patient summaries surface relevant insight from history before visit begins.

Single end-to-end neural network ingesting multi-camera video + navigation + ego-motion, maintaining an evolving 3D occupancy volume as internal world state — spatial memory persisting across frames, tracking objects that temporarily lea…

Agent platform. Keeps a configurable window of recent turns (default 50, range 10–100) in the {vf_memory} variable; older turns are summarized rather than dropped. Chat projects can persist history indefinitely or clear it on tab close/r…

Bi-temporal knowledge graph preserves full lineage of every stored fact — every memory write traceable and auditable by graph schema, not bolted-on logging. SOC 2 Type II; HIPAA BAA available; BYOK + VPC-isolated deployment for data resi…

Distinct from LangMem. Built-in checkpointer saves graph state per super-step (short-term, thread-scoped). Store System adds long-term hierarchical key-value memory across threads with optional vector search + TTL. Postgres / Mongo / Red…

AWS managed memory service for agents running on Bedrock AgentCore. Native integrations with LangChain/LangGraph and Strands. Drives long-running MCP servers + agent workloads.

AutoGen's memory layer: ListMemory chronological context plus teachable agents that vectorise corrections; integrates with Mem0/Zep rather than building deep memory natively. Microsoft AutoGen is in maintenance mode (community managed, b…

Microsoft Semantic Kernel's memory plugin (Microsoft.SemanticKernel.Plugins.Memory): a plug-in approach to long-term memory in the SK agent framework (agent memory marked experimental). Semantic Kernel's README names Microsoft Agent Fram…

Google Cloud managed memory service, now 'Agent Platform Memory Bank' in Gemini Enterprise Agent Platform. LLM-based memory extraction/consolidation from sessions (background, multimodal, custom topics), per-identity isolation, similarit…

Privacy-focused integrated assistant. Memory is user-curated: explicit facts and preferences persist across all sessions, stored locally with no server retention. Agentic browsing experiment (Brave Nightly Dec 2025) takes multi-step acti…

Managed persistent memory service for AI agents on Cloudflare's edge. Private beta announced 2026-04-17. Sits alongside Workers AI / Vectorize as the memory tier of the Cloudflare agent stack.

Systematic literature review + data extraction + meta-analysis. Every extracted claim anchored to its exact source sentence; Notebook persists paper collections across projects. Research Agents (Dec 2025) carry evidence forward across re…

Google's Agent Development Kit. Session services: InMemory / Database / VertexAI. Long-term: InMemoryMemoryService or VertexAIMemoryBankService (uses Vertex AI Memory Bank Preview for persistent cross-session memory).

YC F25 'company brain': connects 50+ sources (Slack, Gmail, Notion, Drive...) into a permission-aware context graph served to agents via API/SDK, returning structured results or LLM-ready summaries; learns from queries over time.

CX support. Currently CRM-backed customer attributes + conversation history lookup. Lifecycle-persistent memory ("priorities learned day one considered in year 10") on roadmap as of Pioneer 2025 — not yet shipped.

First-class memory module inside LangChain — explicit cognitive-science-style typing of memories.

Two-tier model: ChatMessage FIFO queue with overflow flushed into long-term Memory Block objects. Memory class replacing older ChatMemoryBuffer.

Long-term context stored as files in /docs that the agent reads on each session. Agent Mode (default since July 2025) holds working context within session but doesn't persist cross-session beyond filesystem state.

TypeScript-first agent framework. Default memory: LibSQL store + on-device fastembed-js for semantic recall.

Note-taking with no folders; AI-chat over notes. Focused on connections between notes rather than hierarchy.

Renamed NVIDIA CMX Context Memory Storage Platform (2026). A BlueField-4 (STX) powered "AI-native context tier" that extends GPU memory with a shared pod-level KV-cache tier for long-context, multi-turn agentic inference, with Dynamo rou…

OpenAI's agent SDK. Sessions for working context within an agent loop; RunContextWrapper for structured persistent state. April 2026 update added categorised long-term memory: project / user / policy. Long-term tier typically backed by e…

Enterprise voice agents. Context heavily runtime-injected from CRM / telephony / API / user history; passes context to humans on escalation. Strong on plumbing, lighter on autonomous memory accumulation.

Pydantic-AI ships no native persistent memory; agents start fresh each run. Long-term memory typically added via third-party integration (e.g. Hindsight) using tools + auto-injected memory instructions.

Voice agents. Short-term context within call native; long-term memory via webhook integration to CRM / external DB. Cross-agent memory sharing in multi-agent setups.