Streak (no-code algorithmic trading platform)
Streak is a no-code algorithmic trading and strategy automation platform accessible at streak.tech, developed by Streak Tech Private Limited and integrated with Zerodha’s Kite platform via the Kite Connect API. Streak enables retail traders without programming knowledge to create rule-based trading strategies using a visual condition builder, backtest them on historical market data, and deploy them for live automated execution against a Zerodha trading account.
The platform targets the segment of active traders who wish to systematise their trading rules without writing code in Python, Java, or other programming languages. By abstracting strategy logic into a graphical condition interface and connecting live execution to Kite Connect, Streak provides a no-code path to algorithmic trading within the regulatory constraints applicable to Indian retail brokerage.
As of 2024, Streak serves tens of thousands of users across its free and paid subscription tiers. It has become one of the most widely adopted algorithmic trading platforms in the Indian retail market, positioned between completely manual trading (through Kite Web or Kite Mobile) and full-code algo trading (through pykiteconnect or other Kite Connect clients).
History and background
Streak was founded by Praveen Jayaraman and Nikhil Arora, both with backgrounds in quantitative finance and technology. The company launched Streak in 2017 and received investment from Zerodha as part of its ecosystem investment strategy. The partnership with Zerodha is commercial and technical: Zerodha uses Streak’s platform as a preferred no-code strategy tool for Kite clients, and Streak uses Kite Connect for live market data and order execution.
The founding thesis, as described by the founders in interviews and Z-Connect posts, was that retail traders who trade based on technical indicator signals (crossing above/below moving averages, RSI thresholds, candlestick patterns) could benefit from automating the execution of those signals. The gap they identified was between full-code algorithmic trading (which requires Python or similar skills that most retail traders lack) and purely manual trading (which requires constant screen monitoring).
Streak’s no-code visual strategy builder was designed to close this gap: traders define their strategy rules in a form that requires no programming syntax, and Streak’s execution engine translates those rules into market monitoring and order placement actions.
Features
Strategy builder
Streak’s strategy builder is the core product interface. It presents a visual, form-based condition editor where traders define entry and exit rules for a trading strategy.
Condition construction: A single condition is built by:
- Selecting an indicator from the library (moving averages, RSI, MACD, Bollinger Bands, ATR, SuperTrend, Aroon, CCI, Ichimoku, VWAP, volume, candlestick patterns, price-based conditions, time-based conditions, and many others).
- Specifying the indicator parameters (period length, source price, open, high, low, close, VWAP, and any sub-parameters specific to the indicator).
- Choosing a comparison type: greater than (>), less than (<), crosses above (the indicator transitions from below to above the comparison value), crosses below, equals, percentage above, percentage below.
- Specifying the comparison value: a fixed number, the value of another indicator (with its own parameters), or the previous candle’s value for the same or a different indicator.
Combining conditions: Multiple conditions are combined with AND (all must be simultaneously true) or OR (at least one must be true) logic. Complex entries can combine multiple conditions, for example: “RSI(14) crosses above 30 AND price crosses above SMA(50) AND volume > 1.5 * SMA(Volume, 20).”
Entry and exit rule sets: The entry condition set triggers the opening of the position (buy or sell). The exit condition set triggers the closing of the position. Stop-loss and target levels are configurable separately as fixed-price levels, percentage-based levels, or ATR-multiple-based levels (a common risk management approach for sizing stops relative to market volatility).
Trailing stop-loss: Streak supports trailing stop-loss configuration, where the stop price moves in the direction of the trade as the position becomes profitable, locking in gains while allowing the position to continue running.
Candle-close vs intrabar triggers: Streak supports evaluation of conditions at candle close (after the candle period completes) or on an intrabar basis (evaluating live tick data within a candle). Candle-close evaluation is more conservative and less susceptible to intrabar noise; intrabar evaluation allows faster signal capture but may generate more false signals.
Backtesting
The backtesting engine runs the defined strategy against historical OHLCV candle data for the selected instrument and time frame. Backtest inputs include:
- Instrument and exchange
- Candle interval (1-minute, 3-minute, 5-minute, 10-minute, 15-minute, 30-minute, 60-minute, day)
- Date range (from start date to end date)
- Capital allocation (number of units per trade or fixed capital per trade)
- Brokerage and slippage assumptions
Backtest results report:
- Total number of trades
- Win rate (percentage of profitable trades)
- Average profit per winning trade
- Average loss per losing trade
- Profit factor (ratio of gross profit to gross loss)
- Maximum drawdown (largest peak-to-trough decline in portfolio value)
- Equity curve chart (portfolio value over the backtest period)
- Trade-by-trade log with entry and exit prices, P&L, and hold duration
Backtests on paid plans run on minute-level data; the free plan may be limited to daily candle backtesting. Streak’s backtest assumes execution at the signal candle’s close price, which may not be achievable in live trading if market movement is fast. The backtest results are therefore indicative rather than a guarantee of live performance.
Walk-forward and out-of-sample testing: Advanced users can manually split the date range into in-sample (strategy development) and out-of-sample (performance validation) periods to assess whether backtest results are robust or overfitted to the historical data.
Paper trading (virtual deployment)
Before risking real capital, clients can deploy a strategy in paper trading mode. Streak simulates order execution against live market data without submitting orders to the exchange. Paper trading tracks notional P&L, the number of signals generated, and the hypothetical fill prices (assumed as market price at signal generation). Paper trading results diverge from live results due to:
- Slippage: Live orders execute at prices that may differ from the notional fill price if the market moves between signal generation and order submission.
- Liquidity: Low-liquidity instruments may not fill at the quoted price in live trading.
- Order rejection: Live orders may be rejected by the exchange for margin, price band, or other reasons; paper trading does not model rejections.
Paper trading is recommended for validating strategy logic and observing signal frequency before committing to live deployment.
Live deployment
Live deployment connects the strategy to the client’s Zerodha account through a Kite Connect access token. When the strategy’s conditions are satisfied, Streak’s servers submit an order to Zerodha’s OMS via Kite Connect’s order placement endpoint. The order appears in the client’s Kite order book and is subject to all standard exchange rules and margin requirements.
Key operational aspects of live deployment:
Access token refresh: Kite Connect access tokens expire at 07:00 IST each morning. Live strategy deployment requires a valid access token; Streak prompts clients to refresh their token daily, typically by re-logging into Kite from the Streak interface. Streak’s documentation describes automated token renewal mechanisms; clients are advised to ensure token renewal is configured before strategies run at market open.
Multiple strategies: Clients can run multiple strategies simultaneously across different instruments and time frames on a single Zerodha account. Each strategy’s positions appear in the same Kite account; clients must monitor combined margin utilisation across all active strategies.
Position sizing: Position sizes are configured at strategy creation and apply uniformly per signal. Streak does not dynamically size positions based on account equity (fractional Kelly or similar) in its base configuration; fixed unit or fixed capital per trade is the standard approach.
Auto square-off: MIS (intraday) product positions opened by Streak are subject to Zerodha’s standard auto square-off at the broker-defined time before market close (typically 15:10-15:15 IST for equity), regardless of whether Streak’s strategy exit condition has been met.
Scanner
Streak’s scanner feature allows clients to screen the NSE and BSE equity universe for instruments that currently satisfy specified technical conditions. Scanner results are refreshed at candle close intervals and provide a filtered list of instruments matching the condition criteria. The scanner is used for:
- Identifying stocks that have just crossed a moving average (potential breakout candidates)
- Finding stocks with RSI in oversold territory
- Locating instruments where a specific candlestick pattern formed on the most recent candle
- Building a watchlist of stocks meeting fundamental entry criteria (volume spike, price near 52-week high)
Scanner results can be used as a semi-automated watchlist for manual trading or as the basis for defining the instrument universe for a live strategy.
Notifications and monitoring
Streak sends push notifications (via mobile app or browser notifications) and email alerts when:
- A strategy signal is generated
- An order is placed by an active strategy
- A live strategy encounters an error (order rejection, token expiry)
- A paper trading strategy generates a signal (for monitoring without live execution)
The Streak dashboard shows all active deployments with their status (running, paused, completed), the number of signals generated in the current session, the current position status (open or flat), and the realised and unrealised P&L for each active strategy.
Technical architecture
Streak’s monitoring infrastructure consumes real-time market data from Kite Connect’s WebSocket feed. Strategy condition evaluation runs on Streak’s own compute infrastructure, which subscribes to the instruments included in active deployments’ universe. Each candle close (or intrabar tick for intrabar-trigger strategies) triggers a condition evaluation pass for all active strategies monitoring that instrument and interval.
When a condition is met, Streak’s execution engine calls Kite Connect’s order placement endpoint using the client’s stored access token. The order is submitted to Zerodha’s OMS. The response (order ID or error code) is captured and logged.
Streak’s back end must scale horizontally to handle the peak condition evaluation load at 09:15 IST (market open, when strategies start evaluating for the first time in the session) and at common candle boundaries (09:30, 10:00, etc.) when many strategies running on hourly or longer intervals evaluate simultaneously.
Subscription plans
Streak offers free and paid subscription tiers. The free plan has limitations on the number of active strategies, historical backtesting date range, and scanner criteria. Paid plans (Basic, Plus, Premium, and potentially others at different price points) progressively remove these limitations, with the highest tier offering the largest strategy limits, longest backtest history, and minute-level data. Plan pricing is published at streak.tech.
Regulatory considerations
Automated trading through Streak is subject to SEBI’s algorithmic trading regulatory framework. SEBI’s circular SEBI/MRD/DoP/SE/Cir-2/2012 requires that all algorithmic orders be routed through the broker’s OMS and that the broker maintain audit logs. Orders placed by Streak’s servers via Kite Connect satisfy this requirement because they pass through Zerodha’s OMS.
SEBI’s August 2022 consultation paper on algorithmic trading by retail investors proposed requiring brokers to obtain prior approval from exchanges for each algorithm used by retail clients through API access. The practical implications for platforms like Streak were under regulatory discussion as of 2024; Zerodha commented on the consultation paper in a Z-Connect post, noting its concerns about the proposed framework’s impact on retail algo traders.
Streak’s use of Kite Connect for execution means that Streak strategies inherently comply with the requirement that orders pass through the broker’s infrastructure. Strategies that generate excessively high order rates may attract scrutiny under NSE’s and BSE’s circulars on algorithmic and high-frequency trading.
Comparison with competitors
No-code and low-code algorithmic trading platforms in India as of 2024:
AlgoTest: Offers options strategy backtesting with detailed Greeks simulation and live execution across multiple brokers. More options-focused than Streak’s broader technical indicator approach. Strong in the index options strategy community.
Tradetron: A multi-broker automated trading platform with a drag-and-drop strategy editor. Supports more brokers than Streak (including Upstox, Angel One, Fyers, Finvasia, and others) but less deeply integrated with any single broker.
TradingView webhooks with Zerodha middleman services: Some traders use TradingView Pine Script alerts firing via webhooks to a middleware server that then calls Kite Connect for order placement. This approach requires more technical setup than Streak but offers the full flexibility of Pine Script strategy coding.
Python-based custom algo frameworks (pykiteconnect): For developers comfortable with Python, building directly on pykiteconnect provides maximum flexibility. Streak abstracts the programming complexity at the cost of some flexibility in strategy logic.
Streak’s competitive advantages are its deep integration with Kite Connect (Zerodha-invested partner with preferred placement in Kite’s interface), the quality of its backtesting engine, and the breadth of its indicator and pattern library, which covers most retail technical analysis strategies without requiring custom code.
Backtesting methodology and statistical validity
Streak’s backtesting engine evaluates a strategy’s defined conditions against historical OHLCV candle data sourced from Kite Connect’s historical data API. The engine applies the strategy’s buy and sell conditions sequentially across the historical period and computes a P&L series based on simulated entries and exits at the open price of the candle following the signal candle (to approximate realistic execution timing).
Backtesting results on Streak, as with any candlestick-based backtesting engine, are subject to standard caveats that serious users must account for:
Look-ahead bias: If entry conditions are evaluated using the same candle’s close price as the execution price, the system implies entry at a price that was only known after the fact. Streak’s candle-close vs. intrabar execution mode distinction is intended to address this, but users must verify the execution assumption matches their intended real-world behaviour.
Survivorship bias: Backtesting across historical NSE data may include companies that were delisted during the backtest period. If delisted companies are excluded from the historical data universe, the backtest overstates performance by missing losses on positions that could not be exited normally.
Transaction costs: Streak’s backtesting can incorporate brokerage costs, STT, and other charges in its P&L calculation. Users who do not include realistic transaction costs in their backtests will overestimate net returns, particularly for high-frequency strategies.
In-sample vs. out-of-sample performance: Optimising strategy parameters on the same historical data used to evaluate the strategy (in-sample overfitting) will produce backtested results that do not generalise to future data. Streak’s walk-forward testing feature, which tests a strategy on a held-out period after parameter selection, is the recommended approach to assess genuine predictive value.
Users who approach Streak with an understanding of these limitations use the backtesting results as a screening tool for plausible strategies rather than as a precise prediction of future P&L.
See also
References
- Streak. “Streak, no-code algo trading platform”. streak.tech. Accessed May 2026.
- Zerodha Z-Connect Blog. “Introducing Streak, algo trading for everyone”. z-connect.zerodha.com. Accessed May 2026.
- SEBI. “Circular on algorithmic trading”. SEBI/MRD/DoP/SE/Cir-2/2012. March 2012.
- SEBI. “Consultation paper on algorithmic trading by retail investors”. sebi.gov.in. August 2022.
- Zerodha Z-Connect Blog. “Our response to SEBI’s algo trading consultation paper”. z-connect.zerodha.com. 2022.
- Kite Connect. “API documentation”. kite.trade/docs/connect/. Accessed May 2026.