MyAlgoGuru Help
Education Glossary Open the app
Help/Reference/Strategy

Strategy Deployment & Execution Engine

Turn a validated strategy into a live, monitored trading process: bind it to a broker account, start and stop it safely, and watch every signal it generates flow through Risk and Order Management before any order reaches the market.

Where you will find this

Deploy strategy (paper vs live, approval step), Runtime monitor (deployments & signals), Kill-switch / risk controls, Execution history.

Purpose

A strategy on its own only describes behaviour. Deployment brings it to life — you bind an Active strategy to a broker account and a runtime configuration, then start it when you are ready. From that point the engine evaluates the market and generates signals, but it never talks to a broker directly.

The controlled path to market

Every order a deployed strategy generates is submitted through Order Management, where the Risk engine validates it before anything reaches the exchange. The execution engine cannot bypass Risk or place orders itself — a platform invariant aligned with the SEBI/NSE retail-algo framework.

The Strategy Deployment & Execution Engine is responsible for deploying, executing, monitoring and controlling algorithmic trading strategies.

It transforms a configured strategy into a live trading process capable of generating trading decisions and initiating orders through the Order Management module.

The module provides the runtime environment for all automated trading activities.

Objectives

The module will:

  • Deploy trading strategies.
  • Execute strategies continuously during eligible trading sessions.
  • Monitor strategy execution.
  • Generate trading signals.
  • Submit trading requests to the Order Management module.
  • Manage the runtime lifecycle of deployed strategies.
  • Support simultaneous execution of multiple strategies.

Functional Scope

The module includes:

  1. Strategy Deployment
  2. Runtime Configuration
  3. Strategy Start
  4. Strategy Stop
  5. Strategy Pause
  6. Strategy Resume
  7. Signal Generation
  8. Execution Monitoring
  9. Runtime Status Management
  10. Execution History

Runtime Architecture

The Strategy Deployment & Execution Engine is responsible only for runtime behaviour.

It will consume strategy definitions from the Strategy Management module.

It does not modify strategy definitions.

Trading orders generated during execution will always be submitted through the Order Management module.

Strategy Deployment

You can deploy eligible strategies.

Deployment will associate a strategy with its execution configuration.

Deployment does not automatically start execution.

A deployed strategy remains available until undeployed.

Deployment Configuration

Deployment configuration may include:

  • Broker Account
  • Trading Product
  • Trading Capital
  • Position Sizing
  • Deployment Schedule
  • Trading Session
  • Runtime Parameters

Deployment configuration remains independent of the underlying strategy definition.

Strategy Lifecycle

Every deployed strategy will maintain a runtime lifecycle.

Typical lifecycle stages include:

  • Deployed
  • Ready
  • Running
  • Paused
  • Stopped
  • Completed
  • Failed

The runtime lifecycle is maintained independently of the strategy lifecycle.

Strategy Start

You can start deployed strategies.

Only strategies meeting deployment requirements is eligible to start.

Successful startup will initiate runtime execution.

Strategy Stop

You can stop running strategies.

Stopping a strategy will terminate further signal generation.

Stopping does not modify completed trading activity.

Strategy Pause

Running strategies may be paused.

While paused:

  • No new signals is generated.
  • Existing trading records remains available.
  • Monitoring will continue.

Strategy Resume

Paused strategies may be resumed.

Execution will continue using the current deployment configuration.

Signal Generation

During execution the engine will evaluate market conditions according to the deployed strategy.

Signal generation may produce:

  1. Buy Signal
  2. Sell Signal
  3. Exit Signal
  4. No Action

Signal generation does not directly communicate with brokers.

Order Generation

Eligible trading signals will generate order requests.

Order requests is submitted exclusively through the Order Management module.

The execution engine does not communicate directly with broker integrations.

Multi-Strategy Execution

The platform will support simultaneous execution of multiple deployed strategies.

Each strategy will execute independently.

Failure of one strategy does not affect the execution of other deployed strategies.

Execution Monitoring

The platform will continuously monitor runtime execution.

Monitoring includes:

  • Runtime Status
  • Signal Generation
  • Processing Activity
  • Runtime Errors
  • Order Submission Status

Execution monitoring remains available throughout the deployment lifecycle.

Execution Status

Every deployed strategy will maintain an execution status.

Typical statuses include:

  • Idle
  • Initialising
  • Running
  • Paused
  • Waiting
  • Stopped
  • Completed
  • Failed

Execution status will accurately reflect runtime activity.

Execution History

Historical execution information is preserved.

Execution history may include:

  • Start Time
  • Stop Time
  • Runtime Duration
  • Signals Generated
  • Orders Submitted
  • Runtime Events

Historical records remains immutable.

Failure Handling

Execution failures is isolated to the affected deployment.

Failure of one deployed strategy does not interrupt:

  • Other deployed strategies
  • Manual Trading
  • Webhook Execution
  • Order Management

Failures is recorded for operational review.

Business Rules

  • Only Active strategies may be deployed.
  • Only Deployed strategies may be started.
  • Runtime configuration remains separate from strategy definition.
  • Trading signals does not bypass Order Management.
  • Execution does not communicate directly with brokers.
  • Multiple strategies will execute independently.
  • Runtime failures remains isolated.
  • Execution history is permanently preserved.

Audit Requirements

The platform will maintain an audit trail for runtime activities.

Audit events includes:

  • Deployment
  • Undeployment
  • Start
  • Stop
  • Pause
  • Resume
  • Runtime Failure
  • Runtime Recovery

Audit records remains immutable.

Out of Scope

The following are outside the scope of this module:

  • Strategy Design
  • Strategy Versioning
  • Broker Connectivity
  • Order Lifecycle Management
  • Position Management
  • Portfolio Analytics
  • Historical Backtesting

These are covered by their respective business modules.

Running Algos — what is live right now

Running Algos is the live view: which deployments are running, what each is holding, what it has made, and the controls to stop it. What an algo holds is derived from its own orders rather than a separate ledger, so the position shown is the position its fills imply.

  • One deployment runs against one account. A strategy deployed to several accounts is sibling deployments sharing a strategy group.
  • A card’s P&L is an aggregate across legs. A sum that is missing an unpriced leg is flagged as partial rather than presented as the basket’s P&L, and a value that is genuinely unknown reads as a dash rather than zero.
  • A live algo is never archived, so an empty archive is not evidence that nothing ran.

Target, stop and trailing are per leg

Risk controls are evaluated per leg, not across the basket. Each leg measures against its own entry, and a futures leg measures from the price it was actually entered at rather than a reference derived later.

  • Profit Protection is a trailing mode rather than a separate control. There is one ratchet, and where conditions overlap the tightest one wins.
  • A basket P&L that cannot price every leg is not the basket’s P&L. It is reported as partial, because a sum missing a leg is a smaller number than the truth and reads as a safer position than you hold.

Priority Execution

Priority Execution submits a basket’s legs in a declared order with a fill timeout, rather than firing them all at once. It is for structures where one leg should be established before the next is attempted.

Option Pair Scalping

Option Pair Scalping works a ladder across a pair of option legs, holding its state across the pair rather than treating each leg as an independent position. It is an execution engine of its own, so what it is doing is read from its own state rather than inferred from the order book.

A running ladder is trading

Describing what it is doing is safe. Starting or stopping it is your own action.

Square off: one label, two paths

Squaring off a paper deployment closes the simulated position; squaring off a live one sends orders to the broker. The control reads the same in both cases, so which account the deployment is on is the thing to establish first.

Deploy and run a strategy

  1. Confirm the strategy is Active — only Active strategies are eligible to deploy.
  2. Create a deployment: pick the strategy version, the broker account, and runtime config (product, capital, position sizing, schedule).
  3. The strategy is now Deployed but idle — deployment never auto-starts trading.
  4. Press Start when ready. The engine opens an execution run and begins evaluating signals.
  5. Use Pause to hold (no new signals; monitoring continues) and Resume to carry on.
  6. Press Stop to end the run. Completed trades are preserved; the deployment can be undeployed once stopped.

Safety & manual intervention

A running deployment can be paused or stopped at any moment, and the kill-switch halts execution immediately across deployments. Failure of one deployment is isolated — it never affects other strategies, manual trading, or order management.

AI never trades on its own

MyAlgoGuru never executes a trade or other irreversible action without your explicit approval. Automated execution stays bounded by your Risk policies and can be halted instantly.