Original research and field notes on the records organisations keep, the reasoning they lose, and how legal and real estate teams can build a memory that remains governed.
§ 01Featured field note
Organisational intelligence
Field noteSearch finds documents. Organisational memory preserves the decisions, permissions, and provenance that make those documents useful.
Source.
The record that was created
Reasoning.
The decision it shaped
Memory.
The context that survives
Scope attached / source retained / time recorded
§ 02Research & field notes
Real estate
A deal produces years of reasoning, but most of it disappears into folders once the transaction closes. The hold period deserves a living record.
Legal
Firm-wide intelligence only works when every answer respects the same client, matter, and ethical boundaries as the source material.
§ 04Technical library
Earlier explainers, comparisons, and guides on AI agent memory — the technical foundations beneath organisational intelligence.
Comparison
An honest comparison of three AI agent memory layers — Mem0, Zep, and OctaMem — across memory model, MCP support, auditability, and pricing.
Explainer
What semantic, episodic, procedural, and working memory mean for AI agents — with plain-language definitions, examples, and when to use each.
Guide
A step-by-step guide to giving Cursor and Claude long-term memory across sessions using OctaMem's remote MCP server — no local install required.
Research at OctaMem
Evidence for how organisations remember and decide.
We publish applied research, field observations, and technical papers on governed organisational intelligence in legal and real estate work.
§ 41Guides
Cursor permissions.json and mcpAllowlist: The Complete ReferenceHow permissions.json works in Cursor: locations, precedence, mcpAllowlist wildcard syntax, and the three failure modes that silently break your allowlist.Read the guide
What Is an AI Memory Layer?An AI memory layer gives agents persistent state across sessions. How it differs from RAG and vector databases, and seven criteria for choosing one.Read the guide
CoALA Explained: Working, Episodic, Semantic and Procedural MemoryA practical breakdown of CoALA: working, episodic, semantic and procedural memory, the decision loop, and where the framework falls short.Read the guide
How Much Does an AI Memory Layer Cost?What AI memory infrastructure actually costs: the five pricing models, which one hides costs at scale, and the enterprise line items nobody quotes upfront.Read the guide
Semantic Memory in AI AgentsSemantic memory is an agent's store of facts. How it differs from episodic memory, why contradiction handling is the hard part, and how supersession works.Read the guide
Episodic Memory in AI AgentsEpisodic memory lets an agent learn from its own history. Why similarity search is the wrong retrieval strategy for events, and when to consolidate.Read the guide
Procedural Memory in AI AgentsProcedural memory stores knowing-how, not knowing-that. Explicit skills versus weights, why self-written tools are risky, and how to version them.Read the guide
Agent Memory Benchmarks Explained: LoCoMo, LongMemEval and BEAMWhat LoCoMo, LongMemEval and BEAM actually measure, what none of them measure, and how to read a vendor's published benchmark score without being misled.Read the guide
How to Choose an AI Memory Layer: A Buyer's ChecklistTen tests that separate memory layers, each runnable in under an hour on a trial account: contradiction handling, deletion, abstention, latency, export.Read the guide
Nine Ways Agent Memory Fails in ProductionThe memory failure modes that never appear in demos: stale fact confidence, similarity drift, consolidation laundering, and six more, each with a fix.Read the guide
GDPR Article 17 and AI Memory: Why the Right to Erasure Is Architecturally HardMost AI memory systems cannot honour a deletion request completely. Why consolidation breaks erasure, and what Article 17 actually requires.Read the guide
MCP Memory Servers: How They Work and How to Choose OneHow an MCP memory server gives Claude, Cursor and other clients persistent memory: the tools it exposes, local versus remote, and what to check first.Read the guide
Context Engineering vs Memory: What Is the Difference?Context engineering assembles the best prompt for one call. Memory decides what exists to assemble from. Where they overlap, and why you need both.Read the guide
Why Law Firms Need Organisational Intelligence: A Universal Brain for Legal WorkLaw firms run on institutional memory, and it leaves with every departure. Why legal AI without organisational intelligence is an intern with amnesia.Read the guide
Why Real Estate Firms Need a Universal Brain: Organisational Intelligence for PropertyIn real estate, deal and relationship memory produces the revenue, and it lives in agents' heads. What changes when an agency governs it centrally.Read the guide
Organisational Intelligence: Why Every Company Needs a Universal BrainOrganisational intelligence makes what a company knows available to every person and agent, under governance. What it is and how to build it incrementally.Read the guide
Why Your Startup Needs a Company Brain Before Its Next HireA startup's context lives in two founders' heads and evaporates each session. Why the cheapest headcount you can add is a memory layer.Read the guide
A Second Brain for Your AI Agents: Personal Memory That Travels With YouYou built a second brain for yourself. Your agents cannot read it, and their memories are locked per vendor. Why you need one portable memory layer.Read the guide
The best enterprise AI memory platforms in 2026We compared 13 enterprise AI memory platforms, sorted into memory layers, cloud agent services and databases, on deployment, compliance and pricing.Read the guide
Why does my AI keep forgetting? A plain English guide to AI memoryChatGPT, Claude and Gemini forget for the same three reasons. What AI memory really is, why the built-in fixes fall short, and what a company needs.Read the guide
What is a company brain? The plain English guide to organisational intelligenceA company brain is one governed memory of everything your business has learned, which every person and AI agent can ask. What it is, and what it is not.Read the guide
Can law firms use ChatGPT safely? What confidentiality actually requires in 2026Most lawyers already use AI; most firms have no policy. The three real risks of ChatGPT in a law firm, and what a defensible setup looks like.Read the guide
When a partner leaves, what leaves with them? Institutional memory in law firmsEvery firm loses years of judgement when a senior lawyer leaves. What actually walks out, why knowledge management never fixed it, and what does.Read the guide
AI for real estate agents and brokerages: why 82% use it and only 17% see resultsMost agents use AI. Most say it made no difference. What AI really saves time on in a brokerage, what it fails at, and how firms are fixing the gap.Read the guide
Why every property deal starts from zero, and how real estate groups are fixing itLender terms, diligence findings and the reasons behind an investment get lost between acquisition and sale. How to give an asset a memory that lasts.Read the guide
How to start an AI native company in 2026: the founder's playbookAI native founders build the company's memory before they build headcount. The seven moves behind $2M per person, and the stack to start with.Read the guide
How to start an AI native real estate company in 2026 (and why the same plan works for a law firm)Three models for an AI native property business, the stack each one needs, what regulators and clients will ask, and a 90 day plan to launch.Read the guide
The best tools to build an AI startup in 2026: the full stack, comparedThe 2026 AI startup stack, one tool per layer, from model to payments — an honest comparison, plus the layer almost everyone forgets.Read the guide
Every founder should be able to ask their codebase questions, even if they cannot read codeYour codebase is the most honest record of what your company built and why. Why founders who can query it in plain English make better decisions.Read the guide
How to connect GitHub to an AI memory and ask your code, commits and decisions in plain EnglishGive Claude, ChatGPT and your coding agent a shared memory of your GitHub repos: what gets read, how access works, and 20 questions for day one.Read the guide
How to onboard a developer or contractor in a day: give them the codebase's memoryDeveloper onboarding takes weeks because the knowledge is in people's heads. How founders cut it to a day with a memory of the codebase.Read the guide
Claude Code and Cursor forget your codebase every session. The fix also helps your CEO.Coding agents know the code in front of them and nothing about why it is that way. How session indexing differs from durable memory, and the fix.Read the guide
Open source AI is winning. Here is how the best founders build on it in 2026.Open weight models now sit within points of the frontier at a tenth of the price. How founders route across models — and the layer the stack gets wrong.Read the guide
The biggest hole in the open source AI stack is memory, and it matters more than the modelOpen models caught up with the frontier. Open memory did not. Six ways self-built memory fails in production, and what a governed layer does instead.Read the guide
Shared AI memory for enterprise teamsWhat shared AI memory is, how a company owns it rather than an employee, and what changes when someone leaves. The complete guide for enterprise teams.Read the guide
How to share an AI memory with your teamInvite a colleague, grant viewer or editor access to a memory, and understand what each role can actually reach. A step-by-step guide with screenshots.Read the guide
AI memory access control and offboardingWhat actually happens to shared AI memory when an employee leaves — grants, sessions, API key rotation — and the questions to ask any memory vendor.Read the guide
How pooled billing works for shared AI memoryOne company plan, pooled memory operations and storage, no seat counting. How shared AI memory is billed and why the company — not the admin — is the customer.Read the guide
Audit logs for shared AI memoryPer-user activity logs for shared AI memory — which colleague ran which operation, on which memory, what it cost, and how long the records are kept.Read the guide
