NF.

Active Build · System 01

ORION

Autonomous Trading Intelligence

Market conditions become disciplined signals, risk decisions, and execution.

01 / Why

Why ORION
exists.

Trading ideas were available. Consistent execution was not.

Observe the same conditions. Apply the same safeguards. Execute the same process.

02 / Problem

The failure
chain.

Small execution gaps compound. The system was designed to break that chain before a trade reaches the market.

01Manual Chart Monitoring
02Late or Missed Signals
03Inconsistent Decisions
04Weak Risk Discipline
05Unstable Results
03 / Challenges

What had to
work.

Signal quality was only one part of the build. Reliability, risk, control, and simplicity mattered just as much.

Closed-candle monitoring

React only after the market state is complete.

False-signal control

Filter noise and sideways conditions before entry.

Backtest parity

Keep research and live decision rules aligned.

Constrained risk sizing

Size exposure within account and market limits.

Reliable execution

Translate valid decisions into controlled MT5 orders.

Mobile operations

Monitor and control the system through Telegram.

Operational resilience

Detect failures before they become missed entries.

Operable simplicity

Keep the system understandable enough to trust.

04 / Evolution

From idea to
operating system.

Capability was added in layers, with validation between each step.

01Initial Idea
02Manual Signal Logic
03Semi-Automated Execution
04Risk Engine
05Multi-Account Architecture
06Autonomous System
07Continuous Backtesting
05 / Workflow

How ORION
works.

Nine understandable decisions connect market observation to monitored execution.

Market Data
Candle Close Detection
EMA 8/13/21 Trend Check
Touch & Rejection Validation
Sideways & Safety Filters
Risk & Lot Calculation
Order Execution
Telegram Monitoring
Performance Logging
06 / Principles

Designed for
discipline.

The architecture follows a small set of rules that keep automation accountable.

Closed-Candle Decisions

Act on confirmed information.

Rule-Based Execution

Make conditions explicit.

Risk First

Define exposure before action.

Strategy / Backtest Parity

Protect decision consistency.

Mobile Operations

Keep control accessible.

Simple Architecture

Reduce unnecessary failure points.

Fail-Safe Controls

Stop safely when conditions break.

Continuous Validation

Test before changing production.

07 / Direction

What failed.
What stayed.

Progress came from removing complexity as often as adding capability.

Did not work

  • Over-complex strategy variants
  • Unstable stop-loss experiments
  • Constraints that blocked valid trades
  • Time-based exits that weakened outcomes
  • Production defects that interrupted execution

Final direction

  • One clear EMA 8/13/21 strategy
  • Structural stop-loss logic
  • Fixed risk controls
  • Closed-candle validation
  • Simple Telegram operations
  • Separate production fixes and research
  • Automated checks before deployment
08 / Architecture

The current
system.

Distinct components keep data, decisions, risk, execution, records, and operations legible.

MT5
ORION Runtime
Strategy Engine
Risk Engine
Execution Engine
SQLite
Telegram
GitHub Deployment Workflow
09 / Status

Built now.
Designed forward.

Roadmap items are directions—not completed capabilities or commercial claims.

Now

  • Stable production logic
  • Automated signal and execution flow
  • Telegram monitoring
  • Independent risk controls
  • Backtest and live logic alignment

Next

Roadmap
  • Improve market regime detection
  • Improve monitoring and diagnostics
  • Expand controlled strategy testing
  • Improve operational resilience

Long Term

Roadmap
  • Productized trading platform
  • Multi-account management
  • User-friendly setup
  • Modular strategy engine
  • Commercial or subscription model
10 / Lessons

What the build
made clear.

The strongest lessons extend beyond trading systems.

01

Simple systems are easier to trust and operate.

02

Strategy changes must stay separate from production fixes.

03

Backtest parity matters more than adding more features.

04

Risk control is part of the strategy, not an add-on.

05

Operational reliability matters as much as signal quality.

06

Automation should reduce decision errors—not only save time.

Nana Labs · ORION

More than a trading bot: a disciplined operating system for market decisions.

Strategy, risk, execution, monitoring, and review—one repeatable process.

Back to LabsDiscuss This Project

Continue exploring

What comes
next.

Future Labs remain private until there is a useful system story to share.

Coming Soon

Sentinel Risk Tools

Risk visibility and operating controls.

Coming Soon

Operations Intelligence

Decision support for complex operations.

Coming Soon

AI Review Framework

Human-centered automation for structured review.