21+ AI Agents
Active Alpaca Demo
Python/PHP Tech Stack
Learning AI Mode

The Problem

Markets move in milliseconds. Signals like new patents, FDA approvals, insider trades, or geopolitical impulses are scattered across hundreds of channels. A human cannot correlate these "Boom Signals" in time. Whoever connects the information first, wins.

  • Information overload causes blindness
  • Signals are scattered across too many sources
  • Manual analysis is too slow
  • Emotional decision errors in trading

The Solution

The current approach is a swarm of specialized AI agents working together like an intelligence agency. Scouts scan patent databases, news feeds, social media, and government contracts. The "Head Team" (ASLAN, WATCHER, BIG MAMA, LISA) evaluates impulses in an automated team meeting. Many parts are still evolving: I continuously optimize the agents, flows, and decision logic.

Agent Swarm

TA4U is intentionally built as a learning project: I am currently testing 21+ specialized agents across four layers to understand and implement swarm agentics in practice. Scouts collect raw signals, analysis agents compress patterns, the head team prioritizes, and guardrail agents secure risk, logging, and execution.

4

Head Team

ASLAN, WATCHER, BIG MAMA, and LISA form the decision round. They evaluate opportunity, risk, capital, and timing.

8

Scout Agents

Sector scouts for tech, biotech, defense, macro, patents, insider filings, social, and breaking news.

6

Analysis Agents

Condense sentiment, price structure, source quality, and signal strength into actionable hypotheses.

3+

Execution & Guardrails

Paper trading, alerting, logging, and risk checks accompany every step up to demo execution.

My Role

Founder & Architect

Developing Multi-Agent Logic Integration of Alternative Data Sources Strategy Design & Risk Management

Features

Alternative Data Mining

Scanning patents, job postings, and insider filings for true alpha.

Multi-Agent Consensus

No decision without a team meeting: Different perspectives (risk, opportunity, capital) are united.

Multi-Markt Abdeckung

Monitoring US markets (NASDAQ, NYSE) and the Turkish market (BIST).

Boom Effect Detector

Correlation of weak signals across multiple channels into a strong buy signal.

AI Risk Management

BIG MAMA protects the portfolio from emotional overreactions and calculates every position precisely.

Fully Automated Pipeline

From raw impulse to analysis to paper-trading order – all without human intervention.

Website Modules

Included

Head Agents

ASLAN, WATCHER, BIG MAMA, and LISA as the decision layer.

Included

Sector Scouts

Specialized collectors for Tech, Biotech, and Defense.

Included

Impulse Database

Centralized storage and networking of all market signals.

Included

Telegram Alerts

Real-time reporting of AI meetings and decisions.

Results

Manual
Autonomous
Analysis
Single Indicator
Correlation
Strategy
Gut Feeling
AI Consensus
Decision

TA4U is my personal learning and moonshot project in AI and finance. The current focus is not perfection yet, but understanding agents and swarm agentics properly, implementing them in practice, and improving them iteratively. I am still at the beginning, learning with every step, and continuously refining the agents, prompts, and workflows.

Screenshots

What I Learned

What this project is teaching me right now: agents and swarm agentics may look spectacular from the outside, but in practice they start with solid logic, clean data flows, and a lot of iteration. I am still at the beginning, learning every day, and improving step by step instead of pretending the system is already finished.