Overview & Objectives
Customer Overview
Our client is a Gujarat-based proprietary trading firm whose quant team trades index and stock derivatives with intraday and positional strategies. Each strategy ran as its own script with its own broker connection, so there was no single view of positions, risk or performance, and moving a strategy from research to live trading meant rewriting it.
Business Objectives & Challenges
The desk needed one engine to run every strategy: the same code for backtesting and live trading, a central order management system, firm-wide pre-trade risk checks, and a console where traders and risk managers could see and control everything in real time.
Research-to-Production Gap
Strategies were written one way for research and rewritten for live trading, which introduced bugs and delays.
No Central Risk View
Exposure, losses and order rates were tracked per script, with no firm-wide limits.
Order State Management
Orders, modifications, partial fills and rejections had to be tracked reliably across many strategies.
Position Reconciliation
Internal positions had to match broker records at all times, with mismatches flagged immediately.
Market Data at Scale
Tick data for hundreds of instruments had to be captured, stored and replayed for testing.
Low-Latency Signals
Signal-to-order latency had to stay low and consistent during volatile periods.
Operational Control
Traders needed to start, pause or flatten any strategy, and risk managers needed a firm-wide kill switch.
Audit & Reporting
Every decision and order needed a clear trail for internal review and end-of-day reporting.
Solutions
Solutions That Worked
We designed an event-driven engine in Python with a strategy SDK built around on_tick, on_bar and on_order callbacks. The same strategy code runs in the backtester, in paper trading and in live trading, so what the quant team tests is exactly what goes to market. A market data service captures ticks into a time-series database and can replay any session for testing.
All orders flow through a central order management system with a full order state machine, automatic retries and continuous reconciliation against broker positions. Before any order leaves the system, a pre-trade risk layer checks strategy, trader and desk limits: daily loss, gross exposure, order rate and quantity freeze limits. A React operations console shows live P&L, positions, risk usage, latency and an event log, with controls to pause, flatten or kill strategies.
Feature Highlights
Python Strategy SDK
Write strategies once with simple callbacks and run them unchanged in backtest, paper and live modes.
Event-Driven Backtester
Replay historical ticks and bars through the same engine used in production, with realistic fills and costs.
Order Management System
Central order routing with a full state machine, retries, and handling of partial fills and rejections.
Continuous Reconciliation
Automatic matching of internal positions with broker records, with alerts on any mismatch.
Pre-Trade Risk Engine
Loss, exposure, order-rate and quantity checks at strategy, trader and desk level before every order.
Tick Data Capture & Replay
Capture and store market data in a time-series database and replay any session on demand.
Operations Console
Live desk P&L, strategy status, risk usage, latency histograms and event logs in one screen.
Strategy Controls & Kill Switch
Start, pause or flatten any strategy, plus a firm-wide kill switch for risk managers.
Latency Monitoring
Signal-to-acknowledgement latency tracked per order, with alerts when thresholds are breached.
Role-Based Access
Separate permissions for quants, traders and risk managers, with every action logged.
End-of-Day Reports
Automatic reports on P&L, trades, slippage and risk usage for each strategy and trader.
Technologies
Technology Stack
Testimonials
Trusted By Our Clients
Blog
Latest From Our Blog
Looking for a Custom
Application or IT Solution?
Send us your requirements on info@accelfintech.com or call +91 992 513 3221
