Calibrate
Start calibrating
AI strategy backtester

Describe your rules.
Calibrate your edge.

Say what you want from the market. We’ll calibrate the rest: the missing rules, the logic, and a full backtest on a decade of Indian market data.

Build your first strategy See how it works
Calibrate · New strategy
Calibrate · building your strategy
  • ✓Reading your idea and drafting the strategy
  • ?Two choices still need your answer: position size and exit
Calibrate has a quick question How should positions be sized? Equal weight, signal weighted, minimum variance, or describe your own rule.
Equal weight20% of capital per tradeSignal weighted
Strategy ready · HindunilvrRsiEmaLong Runs, finds entries, fills orders and closes positions. Backtesting the last year now.
Net P&L₹ —
Sharpe—
Max drawdown— %
Win rate— %

Illustrative layout. Your own strategy, your own numbers.

2015→ todayA decade of Indian market history
1 minto monthlyCandles of every size, built the standard way
F&O+ stocksEvery listed stock, index, future and option
How it works

One sentence in. A full backtest out.

01

Describe it

Write your strategy the way you’d say it out loud. Instrument, entry, exit, timeframe.

02

We fill the gaps

Calibrate asks about sizing, exits and anything else you left out. No silent assumptions.

03

Get the truth

It checks the logic, runs the backtest and shows every trade, net of costs.

What’s inside

A quant desk that speaks your language.

Broken logic, caught before it runs

Every strategy is checked against a fixed rulebook before a single candle is read, so you never wait on something that can’t trade.

  • Checked against a fixed rulebook
  • Problems explained in plain English
  • Fixed without changing what it trades
  • Every indicator declared and calculated one standard way

Ask why

Question your backtest down to a single candle, then change the rule in the same conversation.

Where are the losses coming from?Why no trades last month?Change the stop loss to 2%

Every trade, on the chart

Entries, exits, P&L and drawdown, candle by candle. Export it all to Excel.

Every number, editable

Plain-English settings. Change one, re-run.

EMA length 21
Candle size 5 min

Your costs, your words

Results after brokerage, the way your broker charges.

₹20 per order plus STT and exchange fees

Fundamentals and technicals together

Filter or rank stocks on valuation, profitability, growth, debt, dividends and promoter holding, then time the entry with RSI, moving averages or a breakout. About 400 company figures, always read as they were known on the day, so a backtest never peeks at the future.

P/E under 20ROE above 15%Debt to equity under 1RSI crosses above 30

Backtest only

Calibrate never places trades or touches your broker account. It’s where ideas get tested, not where money moves.

Machine learning

Go beyond rules. Teach a model.

Ask a question in plain English. Calibrate trains a model to answer it.

Every day at 9:30, predict whether selling a NIFTY strangle 1% out of the money until 3:15 pm will make money.
Decision treeLogistic regressionRandom forestGradient-boosted trees
Try an idea

If you can say it, you can test it.

Buy the NIFTY ATM call when the 5-min EMA(9) crosses above EMA(21)”

Sell the NIFTY weekly ATM straddle at 9:20 with a 25% stop on each leg”

Hold the 20 Nifty 500 stocks closest to their 52-week high, rebalance weekly”

Short TCS when it closes below its 20-day low”

Buy BANKNIFTY futures on an opening-range breakout above the first 15 minutes”

Buy the NIFTY ATM call when Supertrend flips to buy on 15-min candles”

FAQ

Questions

Do I need to know how to code?

No. Describe the strategy in plain English. Calibrate writes and runs the code, and shows you the logic in plain English.

What can I backtest?

All Indian equities, plus F&O on the relevant stocks and indices, on candles from 1 minute to monthly, with data from 2015. You can also combine company fundamentals with technical indicators.

Does Calibrate place trades for me?

No. It’s a research and backtesting tool. It never connects to your broker or touches your money.

Does a good backtest guarantee future returns?

No. A backtest shows how rules would have performed on past data. Markets change, and past performance does not guarantee future results.

Your next idea deserves a test, not a guess.

Build your first strategy