Ajesh M.

I build automated trading systems that trade the strategy you actually designed, and I backtest them honestly before a single real dollar is at risk. Those who have plan to automate their strategies or those who need a trading application to manage all your trades systematically are welcome. I'm a Data Scientist (4+ years, MTech in Data Science, IIT Palakkad) who now works full-time on algorithmic trading: converting strategies into code, backtesting and optimizing them properly, and running them live on MT5, Tradovate and prop-firm accounts with monitoring that catches problems before they cost money. What I've built recently: A multi-bot trading platform: A live execution bridge running many strategy bots across gold, forex, indices and crypto. A FastAPI "brain" filters every signal by market regime (trend/volatility) before execution. The system includes PostgreSQL, full trade and execution audit logs, and a monitoring dashboard. Prop-firm trading apps: Two production apps (demo "monitor" + live prop account "manager") that trade gold ORB, crypto and multi-symbol strategies with exact %-risk position sizing. They include: * Broker-side stop orders * Automatic reconciliation of positions and broker history * Duplicate-order protection * Weekend and outage handling * An analysis dashboard whose backtest history updates itself every 4 hours Futures automation: Strategy research and bots for micro futures (MNQ, MES, MGC, MCL, MBT), using minute-level futures data. Strategy research: I've turned dozens of strategy documents, PDFs and "guru" setups (ORB, SMC / order blocks / FVG, VSA, liquidity sweeps, mean reversion, SMA systems) into exact, testable rules, then measured whether they really work. How I backtest (this is where most bots fail): * 1-minute execution simulation with real bid/ask spread, commissions, slippage and overnight swap * No look-ahead: signals use only closed candles, exactly like live * Walk-forward / out-of-sample validation: tune on one period, prove it on another, and on a market the tuning never saw * Parameter sweeps and robustness checks (RR, filters, timeframes, sessions, volume thresholds), not just one lucky setting * Parity testing: the live bot is checked trade-for-trade against the backtest before it goes live * Clear reports: PDF and dashboard results over 3, 6, 12 and 24 months with win rate, profit factor, drawdown, monthly returns and trade logs I'll tell you honestly when a strategy doesn't hold up. A backtest that flatters a bad idea costs you real money later. AI-assisted trading and monitoring: * Regime filters and signal "brains" that decide which strategy is allowed to trade right now * Automated monitoring: daily/4-hourly health checks, trade reconciliation, discrepancy analysis between accounts or brokers, and alerts * Telegram signal parsing and execution, plus AI-assisted trade validation and reporting I can help you with: * Turning your strategy (PDF, video notes, written rules, TradingView idea) into a working bot * Backtesting and optimizing an existing strategy, including an honest verdict on whether it has an edge * MT5 / MetaAPI, Tradovate and prop-firm bot development with strict risk rules * Fixing a live bot that behaves differently from its backtest * Dashboards, trade journals and automated performance reports Stack: Python, pandas/NumPy, FastAPI, PostgreSQL/SQLite, MetaAPI (MT4/MT5), Tradovate API, Binance data, Docker Happy to start with a small paid task, such as backtesting one of your strategies, so you can judge the quality before a bigger project.