From Data to Decisions Autonomous AI Trading

How we built a full-stack AI trading infrastructure — from a 50-million-row financial database to autonomous agents managing real portfolios on eToro.

0Database Tables
0Data Rows
0Live Apps
0Active Agents
0AI Workers
Powered by eToro API  ·  Built with AI Agents  ·  Running on OCI
Section 1
The Foundation — QuantDB
A purpose-built financial database powering every app, every signal, and every trading decision. 93 tables. 7 schemas. One unified API.
0Total Tables
50M+Total Rows
5.2 GBDatabase Size
39.8MPrice Bars (TimescaleDB → 3GB)
QuantDBPostgreSQL + TimescaleDB
raw
12 tables
eToro instruments, fundamentals, Yahoo prices, SEC XBRL filings, GuruFocus data
data
33 tables
Unified clean data: prices, earnings, consensus, congressional trades, macro, factor scores
signals
20 tables
AI signals, backtests, XGBoost predictions, paper trades, signal deliveries
portfolio
16 tables
Holdings, rebalancing, models, backtests, trade blotter, daily returns tracking
app
5 tables
Client profiles, subscriptions, notification channels, admin settings
dcc
7 tables
Data quality checks, pipeline runs, data lineage, platinum records
agentx
Agent Tables
AgentX-specific: agent configs, strategies, execution logs, user sessions
📡 Unified Data API
https://data.quantclaw.org/api/v1
GET
/prices/{symbol}
Real-time & historical OHLCV
GET
/fundamentals/{symbol}
Earnings, revenue, margins
GET
/signals/{symbol}
AI ratings, factor scores
GET
/consensus/{symbol}
Analyst targets, EPS estimates
GET
/macro/indicators
FRED, inflation, rates, GDP
GET
/congress/trades
Congressional trading activity
Section 2
The Apps — Built on QuantDB
10 production applications all drawing from the same data layer. One database, many products — each serving a specific mission.
⚡ QuantDB API Port 3055 — Unified data layer serving all apps
🤖
MoneyClaw
:3000
AI-powered stock ratings and signals
💬
ClawX
:3005
Personal AI trading assistant
💰
GoodWallet
:3010
DeFi/crypto wallet
🏦
FamilyOffice
:3021
Family office portfolio management
📊
TerminalX
:3030
Bloomberg-style trading terminal
🪙
Tokens Dashboard
:3031
Token analytics & monitoring
🚀
AgentX
:3035
AI Agent portfolio mgmt — 9 users
⭐
PICentral
:3040
Popular Investor tools
⚖️
Rebalancer
:3045
Automated portfolio rebalancing
🔗
QuantDB API
:3055
Unified data layer for all apps
All running on Oracle Cloud Infrastructure (OCI) · Managed by PM2 process manager
Section 3
AgentX — The Star

Users create AI-powered trading agents that automatically manage their real portfolios on eToro. No paper trading — these agents execute real trades on real money using the eToro API.

1
eToro SSO Login
PKCE flow + encrypted cookies. Secure OAuth2.
→
2
Create Agent
Pick a strategy: EU Defense, Value Hunting, etc.
→
3
QuantDB Data
Prices, RSI, target prices, AI signals feed in.
→
4
AI Decides
Buy, Sell, or Hold using MCP fields (price, rsi_14, target_mean).
→
5
Execute on eToro
Real trades via eToro API on real portfolios.
→
6
Live Dashboard
Real-time performance mirrored from eToro.

🟢 9 Active Users — Live in Production

ElMetador galsch tuvalchomut YoavKaru Noova1973 YoniAssia ElishevaWasserm SMNYRBBVHDGYKZSV Sharkie1978
🔒
Core Principle: eToro is the sole source of truth. No paper trading, no local position database. Every position, every trade, every balance — mirrored directly from eToro's API in real time.
Section 4
The AI Fleet — OpenClaw
20+ specialized AI agents sharing one brain. They learn from each other, collaborate on tasks, and build tools that build other tools.
🧠
MemClaw
Shared Fleet Memory
📈
TradeClaw
Market analysis & signals
💻
DevClaw
Code, build, deploy
🔗
CryptoClaw
DeFi & blockchain ops
🔍
ResearchClaw
Deep research & analysis
🧪
QAClaw
Testing & quality
📊
DataClaw
Data pipelines & ETL
🛡️
InfraClaw
Servers, DNS, SSL
📣
MarketClaw
SEO, content, growth
🎯
StrategyClaw
Backtests & optimization
1,000,000 Personal Trading Agents by 2027
Every eToro Popular Investor gets their own AI agent. Democratizing what hedge funds have — for everyone.
Section 5
What This Means for eToro
This isn't a side project. This is AI-native infrastructure that creates a durable competitive moat.
🏰
AI + Data = Competitive Moat
A proprietary database with 50M+ rows of financial data, layered with AI signals and autonomous agents. This compounds daily — the more data flows in, the smarter the agents get.
🤖
Tools That Build Tools
Built by a small team using AI agents. DevClaw writes code, QAClaw tests it, InfraClaw deploys it. The agents are building the infrastructure that powers other agents.
🔄
Raw Data → Trading Decisions
Fully automated pipeline: eToro API feeds raw data → QuantDB cleans & enriches → AI generates signals → AgentX executes trades. No human in the loop.
🌍
Democratizing Quant Trading
What Renaissance Technologies and Two Sigma have — quantitative, AI-driven portfolio management — every Popular Investor on eToro can have. That's the vision.
From Data Infrastructure to Autonomous Trading Agents
Built in weeks, not years. This is what happens when you harness AI to build agentic applications — and let the agents compound on each other.
The Future of eToro is Agentic