·
AI
Stock market: backtesting, strategy automation and algo tradingCase study draft, anonymisedBuild a trading strategy by describing it.
Traders built strategies by picking rules one at a time, from thousands of possible conditions. We built a chatbot instead. Say when to enter and when to exit, and it sets up the conditions for you.
1,000sPossible conditions a strategy can use
1 sentenceTo set both entry and exit
5 key numbersPrice, volume, moving averages, stock on stock, Nifty 50
CustomStrategies made easy, not only standard ones
WhoA backtesting, strategy automation and algo trading platform. Name not published.
BuiltA strategy chatbot that turns plain language into entry and exit conditions
AreaAI: agents and generative AI solutions, AI implementation and automation
From menus to a sentence
Say the strategy. Skip the rule picker.
On the left, a custom strategy built one dropdown at a time. On the right, the same strategy in one sentence, turned into clear entry and exit conditions.
Before: rule pickerRule 10 of ?
EntryExitFiltersRiskAdvanced
Price ▾crosses above ▾SMA 50✕
Volume ▾greater than ▾Avg 20✕
SMA 20 ▾less than ▾Close✕
Stock X ▾correlation ▾Stock Y✕
Nifty 50 ▾% change ▾> 0✕
RSI 14 ▾below ▾30✕
Time ▾between ▾09:20✕
EMA 9 ▾crosses ▾EMA 21✕
Price ▾below ▾SMA 20✕
Stop loss ▾percent ▾1.5✕
+ Add conditionAND / OR ▾1,000s of options
After: strategy chatbotReady
Buy when price crosses above the 50-day average and volume is above its 20-day average, but only if the Nifty 50 is up on the day. Exit when price falls below the 20-day average.
Here is your strategy. Please review it:
Enter whenPrice above 50-day avgVolume above 20-day avgNifty 50 up on the day
Exit whenPrice below 20-day avg
Does this match what you meant? Once you confirm, we will execute it.
Confirm and executeEdit conditions
Confirmed
Strategy confirmed. Sent for execution.
Describe your strategy…↑
The number of signals
Thousands of possible conditions behind every strategy.
Price, volume, moving averages, one stock's impact on another, the Nifty 50, and many more. Each can be combined with the others, which is why picking them one rule at a time stopped working.
Key numbers a condition can useExamples
PriceVolumeMoving averagesStock X on stock YNifty 50RSIEMA crossoversBreakoutsGapsOpen interestVWAPTime of dayBollinger bandsSector indexVolatilityHighs and lows+ thousands more combinations
Highlighted: the five named in this case. The rest show the kind of signals traders combine.1,000sof conditions a custom strategy can draw on
Before and after
From picking rules to saying what you want.
Before
- Entry and exit were set rule by rule in a frontend
- Thousands of conditions to hunt through
- Custom strategies were the hardest to build
- Users had to know where each rule lived
After
- The trader describes the strategy in plain words
- Entry and exit are captured in one conversation
- The chatbot shows the strategy back for confirmation
- Once confirmed, the strategy is executed
The problem
Thousands of conditions, picked one rule at a time.
On the platform, a strategy was built in a frontend. The trader selected the rules to enter a trade and the rules to exit it. But the possible conditions run into the thousands: price, volume, moving averages, one stock's impact on another, the Nifty 50, and more.
What was broken
- Entry and exit were set rule by rule.
- Thousands of possible conditions to choose from.
- Custom strategies, the ones traders really wanted, were the hardest.
You might be here if
- Your users set up their logic by picking rules in a frontend.
- The conditions they might combine run into the thousands.
- Your users need custom setups, not only standard ones.
- You want users to describe what they want, not hunt for the right rule.
The approach
Do not make the rule picker bigger. Let the trader say it.
The chatbot takes the conditions in plain language, entry and exit together. A strategy is gathered in one conversation, not assembled rule by rule.
The trader
A sentence, not a formWhen to enter, when to exit
Built by woodfrog
Strategy chatbotReads the conditions in plain language
Custom strategyEntry and exit conditions
The platform
Backtesting
Strategy automation
Algo trading
Building blocksPriceVolumeMoving averagesStock X on stock YNifty 50and more
The rule picker is still there. The chatbot is a faster way in.schematicThe solution
A chatbot that turns a sentence into a strategy.
What was built
- A strategy chatbot that takes, in plain language, all the conditions needed to enter and exit a position.
- Conditions across the key numbers: price, volume, moving averages, one stock's impact on another, the Nifty 50, and more.
- A custom strategy as the output, with its entry and exit conditions laid out clearly.
- A confirm step: the chatbot shows the strategy back to the trader, and nothing is executed until they confirm it.
The impact
A clean, easy way to build custom strategies.
Traders now describe the strategy they have in mind, instead of hunting for each rule in a menu.
What we will measure with the clientFigures to be confirmed
Time to build a custom strategyBefore and after, the direct proof of easy
Custom strategies builtBefore and after
Conditions the trader confirms as correctThe chat must read the sentence right
Chatbot use against the rule pickerWhich way in people prefer
Let your users say what they want.A short call, or record a message. A person replies within one business day.
