AITrader

An AI-native hedge fund operating system for paper-first trading.

AITrader combines portfolio management, autonomous research agents, configurable trading bots, deterministic risk controls, and broker integrations into a cloud-native SaaS platform built for always-running market intelligence.

Create paper-trading accountSign in

V1 operating stance

Live trading is disabled for V1 until explicitly approved.
AI recommendations stay separated from trade execution.
Risk checks run before execution, not after losses occur.
User-uploaded executable bot code is out of scope for V1.
Paper trading is the default. Live trading remains behind an explicit release gate.

Paper-first Alpaca trading with live-trading release gates

Multi-agent AI hedge fund architecture

Deterministic risk engine before every order

User-selected AI providers and model routing

Subscription SaaS with Stripe billing

Rancher/Kubernetes operations and observability

Platform modules

Built as modular trading infrastructure, not a single black-box bot.

Each module can mature independently: onboarding and broker sync, AI research, strategy bots, risk controls, subscriptions, and operations. The goal is a system that can explain decisions, enforce limits, and scale safely.

Portfolio Command Center

A Robinhood-style portfolio dashboard for account value, cash, buying power, positions, watchlists, order history, and real-time quote context.

  • Portfolio and position views
  • Paper/live environment indicator
  • Performance, drawdown, and allocation panels

Alpaca Broker Integration

Secure broker connection workflows that start with paper trading and keep credential validation, portfolio sync, and order execution auditable.

  • User-configurable Alpaca keys
  • Paper account validation
  • Order and position synchronization

AI Trading Bots

Configurable strategy bots that generate signals, explain trade ideas, and operate under explicit user risk settings rather than free-form automation.

  • Momentum and mean reversion bots
  • ETF rotation and pairs concepts
  • Per-bot risk and model settings

AI Hedge Fund Agent Team

A coordinated agent system modeled after an investment desk: CIO, market regime, risk manager, hedging, execution, and performance analyst roles.

  • CIO Bot strategy coordination
  • Market Regime Bot context
  • Performance Analyst Bot reviews

Risk Manager & Kill Switches

The highest-priority system: deterministic limits that can reject, throttle, or halt AI recommendations before they become broker actions.

  • Max daily loss and drawdown limits
  • Exposure and correlation controls
  • Automated kill switches and stop-loss enforcement

Market Research Agents

Persistent research workers for market news, sentiment, sector rotation, earnings context, and strategy notes that feed human review and bot signals.

  • News and sentiment analysis
  • Regime and volatility summaries
  • Watchlist research briefs

AI Gateway & Model Routing

Centralized model access for OpenAI, Anthropic, Gemini, and local Ollama models with budget controls, failover, and per-bot selection.

  • User-provided API keys
  • Per-bot model selection
  • Cost tracking and fallback routing

Subscriptions & Usage Billing

Stripe-backed subscription packaging for AI features, usage tiers, customer portal access, invoices, and future strategy marketplace monetization.

  • Stripe subscriptions
  • Usage-based billing hooks
  • Customer portal and invoice tracking

Operations & Observability

Kubernetes-native deployment with logs, metrics, traces, CI/CD, secrets management, and runbooks so trading workloads are observable and recoverable.

  • Rancher/Kubernetes Helm deployment
  • Prometheus, Grafana, and Loki
  • Vault/OpenBao-style secrets handling

Agent desk

A coordinated AI investment team with deterministic oversight.

The long-term vision is not one model making unchecked trades. AITrader separates research, regime analysis, signal generation, execution preparation, risk approval, and performance review into distinct roles.

CIO Bot

Dedicated responsibility, auditable outputs, and handoff points into the risk engine.

Market Regime Bot

Dedicated responsibility, auditable outputs, and handoff points into the risk engine.

Risk Manager Bot

Dedicated responsibility, auditable outputs, and handoff points into the risk engine.

Hedging Bot

Dedicated responsibility, auditable outputs, and handoff points into the risk engine.

Execution Bot

Dedicated responsibility, auditable outputs, and handoff points into the risk engine.

Performance Analyst Bot

Dedicated responsibility, auditable outputs, and handoff points into the risk engine.

Roadmap

From safe paper-trading MVP to marketplace-scale strategy platform.

Phase 1 — MVP

Authentication, Alpaca paper trading, core bots, risk manager, dashboard, and safe onboarding.

Phase 2 — Intelligence

AI news analysis, market regime detection, hedging workflows, and stronger multi-model routing.

Phase 3 — Advanced Strategies

Portfolio optimization, volatility strategies, statistical arbitrage concepts, and richer analytics.

Phase 4 — Marketplace

Strategy marketplace, user-created bots, social/institutional tooling, and broader broker support.

Architecture

Next.js frontend, FastAPI services, PostgreSQL, Redis/NATS, optional TimescaleDB, and AI routing through hosted or local providers.

Security

Encrypted broker credentials, audit logging, rate limiting, MFA-ready authentication, and explicit separation of advice from execution.

Runtime

API base: /api; base path: . Built for Rancher/Kubernetes deployment with observable services.

Contact me

Ask about AITrader paper-mode access

Send a message to the Didur Labs contact mailbox. Include your reply email so we can follow up.