DEV Community

Igor Ganapolsky
Igor Ganapolsky

Posted on

AI Trading: Day 89 - 4 Lessons Learned (January 25, 2026)

Day 89/90 - Sunday, January 25, 2026

Every mistake is a lesson in disguise. Today we uncovered a critical flaw in our system - the kind that separates amateur traders from professionals who survive long-term.

4 lessons learned today (1 critical, 1 high priority)

CTO Ignores Surfaced RAG Lessons - Pattern Identified

Every session:

RAG Learning Synthesis - Iron Condor Adjustments

During Ralph Mode iteration 21, queried RAG and synthesized key learnings from recent lessons.

Iron Condor Optimal Control Research

LL-309: Iron Condor Optimal Control Research

Date: 2026-01-25
Category: Research / Strategy Optimization
Source: arXiv:2501.12397 - "Stochastic Optimal Control of Iron Condor Portfolios"

Key Findi

VIX Timing for Iron Condor Entry

LL-310: VIX Timing for Iron Condor Entry

Date: 2026-01-25
Category: Strategy / Entry Timing
Status: RESEARCH

Key Finding: IV Rank and VIX Level Matter

Optimal Entry Conditions

| Parameter | Rec


Tech Stack Behind the Scenes

Our AI trading system uses:

  • Claude Opus 4.5 - Primary reasoning engine for trade decisions
  • OpenRouter - Cost-optimized LLM gateway (DeepSeek, Mistral, Kimi)
  • Vertex AI RAG - Cloud semantic search with 768D embeddings
  • Gemini 2.0 Flash - Retrieval-augmented generation
  • MCP Protocol - Standardized tool integration layer

Every lesson is stored in our RAG corpus, enabling the system to learn from past mistakes and improve continuously.

Full Tech Stack Documentation


Auto-generated from our AI Trading System's RAG knowledge base.

Follow our journey: AI Trading Journey on GitHub

Top comments (0)