Built for the Indian market, not adapted to it

1. Why we built it

Most retail algorithmic trading tools are built for US equities and then translated for India. The translation leaks in the places that decide whether a strategy is actually profitable: lot sizes, circuit limits, the settlement cycle, and above all the cost stack. A backtest that ignores STT, stamp duty, exchange transaction charges, GST and SEBI turnover fees will show you a winning strategy that loses money the moment it meets a real broker.

GTP Algo starts from the Indian cost model rather than bolting it on. Every backtest is priced with the charges a real order would attract, plus a slippage assumption you control, so the equity curve you approve is the one you can expect to trade.

2. How we approach it

Three commitments shape the product, and they are the reason it looks the way it does:

No code required, no ceiling either. Strategies are composed visually — conditions, multi-step flows and candle patterns — because requiring Python excludes most of the people who actually understand a market. But the builder is not a toy: anything expressible in the engine is expressible in the builder.

Honest numbers over flattering ones. It would be trivial to produce prettier backtests by quietly assuming mid-price fills and zero costs. We would rather show you a smaller number you can rely on.

Speed is a feature of thinking, not of benchmarks. Research is iterative. A backtest that takes ten minutes gets run once; one that takes two seconds gets run forty times, and the fortieth version is the good one.

3. The engine

The backtest engine is written in Rust and runs server-side against real NSE and BSE history. That choice is about iteration speed rather than benchmarking pride — sweeping a few hundred symbols across several parameter sets is the difference between a hunch and a result, and it is only worth doing if it returns while you are still thinking about the question.

The web terminal and the Android app are clients of the same engine, so a strategy tested on a laptop and monitored on a phone are the same strategy, not two implementations that drift.

4. Who it is for

People who already understand markets and want to test what they believe. You do not need to write code, but you do need a view worth testing — the product will not supply one, and it deliberately ships no signals, no tips and no managed portfolios.

5. Company details

GTP Algo is operated by FULL LEGAL ENTITY NAME (PROPRIETORSHIP / LLP / PVT LTD), registered in India.

6. Talk to us

Product questions, access requests and anything else go to support@gtpalgo.com. For the right address for privacy requests or a formal grievance, see the contact page.

If you would rather just look at the product, create an account or read how it works.