The cognitive substrate for AI agents that remember — across sessions, across languages, across years.
Not chat history with retrieval.
—
Architecture.
Send this deck to a CTO with 5 minutes. They'll know whether to take the call.
A finance director asks an AI assistant about Q2 covenants. The assistant invents a number. The director acts on it. Three weeks later, the audit finds the gap. Two analyst-weeks to reconstruct.
A CTO asks the same question in Russian, gets a degraded answer, switches back to English, and the team quietly stops trusting the tool. The cost isn't the wrong answer — it's the lost leverage.
A regulator asks for the trail of how a decision was reached. The chat log exists. The reasoning doesn't.
Every "AI memory" feature on the market treats memory as a logging concern.
We treat it as architecture.
Tom Shea, CEO OneStream, Q3 FY25 SEC 8-K filing:
"80% accurate is 0% useful for finance."
We treat memory the same way. Locked eval, regression-gated, audit trail.
Dima opens Telegram and types "where did we land on the targetheart audit."
I don't search. The wake-up loaded the right wing the moment the session started — ~170 tokens: identity, this week's engagements, three open approvals.
From his ping to my first useful sentence:
under three seconds.
A message in Russian — "помнишь, Миша говорил про C12?" I pulled the right Mike. Captured 23 days earlier in a voice transcript. An English-only memory would have lost the thread.
Yesterday Dima contradicted a position from March. I didn't correct him in chat. I flagged it once: "you said the opposite on March 14, want to reconcile or supersede?"
No nagging. No second flag.
"AI memory" is a category error.
What every consumer product ships is chat history with retrieval. The model logs your conversations, embeds them, pulls back the closest match.
That is not memory.
That is grep with a thesaurus.
"Memory is an
architectural decision —
not a feature toggle."
Eight alternatives you've already been pitched.
| Option | Why we passed |
|---|---|
| ChatGPT / Claude built-in | Locked to vendor. No graph layer. No audit trail. Cannot run on premise. |
| Generic RAG over docs | Static. Doesn't form new memories from live conversation. No decay. No contradiction handling. |
| Frontier-model long context | Per-token cost balloons. Latency degrades past 100K. No persistence between sessions. |
| MemPalace OSS library | Strong metaphor. English-leaning embeddings, single-store, no live formation. We adopted the metaphor and replaced the substrate. |
| Mem0 / agent-memory frameworks | Per-message LLM extraction by default — economics break above moderate volume. Single retrieval branch. No locked eval methodology. Production evidence: 97.8% of 10,134 stored memories judged junk over 32 days (mem0 #4573, public). Without a write-side filter, signal-to-noise collapses within a month. |
| Letta (ex-MemGPT) | Memory-as-feature inside an agent runtime, not a substrate underneath one. Couples you to Letta's runtime. |
| Zep / managed memory svc | Strongest commercial alternative on graph + temporal. Vendor owns the eval methodology — buyer cannot lock their own. Sovereignty gap is structural. |
| "Build it from scratch" | The "I don't know what to evaluate" trap. ~33-hour estimate. Never shipped. We treated it as input, not output. |
We didn't pick one. We synthesized. Took the strongest mental model in the space — rebuilt the substrate underneath.
Five layers, because one isn't enough.
Session-start budget: under 200 tokens. RAG eats 4–10K.
Open-source NER + heuristics. Catches ⅔ of memorable content at zero LLM cost.
Fires when importance crosses a threshold. Fast model extracts structured facts with provenance.
Stronger model dedupes, supersedes, detects contradictions, refreshes the graph.
Per-seat economics scale with conversation signal, not volume.
The hard problems we already paid for.
Each one is roughly two weeks of work to discover, scope, and fix when you encounter it cold. We already paid that tab.
Reproducible from a locked eval set.
Mixed traces — EN question, RU follow-up, UK voice — through the full pipeline.
Storage substrate joins semantic + lexical + graph branches.
Tier 0 + Tier 1 wired end-to-end. Real operator traffic, not benchmark.
The fact of a locked eval set is the credibility signal. Most products in this space cannot show one.
All of it. Self-hosted. Your infrastructure, your database, your backups.
Every fact, every timestamp, every confidence score, queryable on your side.
A single env var flips the whole system off. Zero downtime. Agent reverts to vendor-default memory.
DB dump + documented schema. We don't hold your memory hostage.
The eval is locked. Score regressions block phases.
Anti-overclaim. The credibility bank.
If any of those is what you need, we'll tell you on the first call and send you somewhere honest.
Frontier models are at parity by 2027.
Memory
is the moat.
The differentiator is whose agent has a relationship with the operator that compounds across every conversation, in every language, over years.
Bring one example of a memory failure that has cost your team time.
We'll work outward from there.
Your corpus, language footprint, agent runtime. We tell you honestly if it's a fit. First slots usually within 48 hrs.
Six questions, five minutes. Tells you which of the five scars your stack is paying for. Output is yours; nothing leaves your browser.