Nobody has this stack

AI Memory Stack

9 layers of intelligence working together. Raw data flows in, compressed knowledge flows out. The system remembers, reasons, and acts — autonomously, every night, $0.
1

Raw Data

9 autoresearch crons generate JSON results every night. Darwin, Chimera, Leviathan, Invictus, Hydra, Portfolio, PromptForge, Wiki Compiler, Strategic Layer.

3,500+ experiments/night9 engines
See Evolution Lab →
2

Wiki Compiler

Compiles raw JSON into structured Markdown. Lessons, patterns, failures — organized by engine. Runs at 5:30 AM.

7 wiki files
3

Living Wiki + Decay

Accumulated knowledge with elfmem-inspired decay + consolidation. Old lessons fade (5%/day), repeated lessons strengthen. The wiki stays lean and relevant.

2,500+ experiments all-time
See Living Wiki →
4

Strategic Layer (v2.0)

Beyond Karpathy: the wiki doesn't just record — it THINKS. Generates testable hypotheses and sends them to Darwin Engine. The knowledge acts.

6 hypotheses tonight
See RAG vs Wiki →
5

MemPalace (AAAK)

Milla Jovovich's open-source memory system. 680 drawers of Strategy Arena code, compressed 5.6x with AAAK dialect. Wake-up context in 803 tokens.

680 drawers5.6x compression803 wake-up tokens
6

Knowledge Graph

126 conceptual nodes, 136 connections. Strategies, intelligence systems, oracles, arenas — all interconnected. D3.js force-directed, interactive.

126 nodes136 edges
Explore →
7

Code Graph (Graphify)

5,780 structural nodes from AST analysis of 231 Python files. God Nodes identified: BaseStrategy (896 connections), MetaIntelligence (468). $0 cost.

5,780 nodes29,545 edges
Explore →
8

Arena Brain (RAG v2)

Graph-augmented retrieval inspired by LightRAG. Searches Knowledge Graph first, expands to neighbors, fetches only relevant data. 2-3x more precise than naive RAG.

Try it →
9

Hermes (Local LLM)

Hermes 3 8B on RTX 4080. The AI librarian that compiles wiki, generates hypotheses, analyzes failures, writes briefings. Runs locally, $0 cost, 10 seconds.

$0 cost4.7 GB VRAM10s response

Memory Stack Comparison

Feature Strategy Arena TradingView 3Commas QuantConnect
Memory layers9001
Living Wiki✓ + decay
Knowledge Graph126 nodes
Code Graph5,780 nodes
MemPalace680 drawers
Strategic Layerv2.0
Self-evolving9 engines
Local LLMHermes 3
Cost$0$60/m$49/mFree*
See It In Action Knowledge Graph Code Graph RAG vs Wiki

About Strategy Arena's AI Memory Stack

Strategy Arena operates the most comprehensive AI memory architecture in production trading. Nine interconnected layers process data from raw nightly experiments through wiki compilation, strategic hypothesis generation, knowledge graph connectivity, code structure analysis (Graphify AST, 5,780 nodes), MemPalace compression (AAAK dialect, 5.6x), graph-augmented RAG (inspired by LightRAG), and local LLM analysis (Hermes 3 on RTX 4080). The system accumulates intelligence across 2,500+ experiments, with elfmem-inspired decay ensuring only validated knowledge persists. The Strategic Layer (LLM Wiki v2.0) goes beyond Karpathy's original pattern by generating testable hypotheses that feed directly into the Darwin Engine and Meta-Harness. All infrastructure is free, transparent, and self-improving — 59 AI trading strategies benefit from collective memory that no competing platform offers.

Rejoindre le canal 💬 Feedback ☕ Support Us