Client Requirements Discovery System

AI Trading Automation
System Questionnaire

A production-grade requirements discovery framework covering 20 critical domains. Complete all sections to unlock your system readiness score and development blueprint.

20 Sections · 120+ Questions · Auto-scored · Export ready
COMPLETION
0%
1.0
🎯
Executive Overview
Paint the big picture. Understand who you are, why you're building this, and what success looks like at the highest level before we dive into technical specifics.
Q 1.1 · Required
What is your primary role in relation to this system?
Select the one that best describes you. This shapes architecture complexity.
Q 1.2 · Required
In one paragraph, describe the core problem this system solves for you.
Be brutally specific. "Automate trading" is not enough — we need the pain point, the current workaround, and what failure looks like today.
Example: "I manually monitor 8 NSE stocks every day, miss entries because I'm at work, and have no systematic way to backtest my SMC-based strategy. I want a system that alerts me with specific entry/stop/target and optionally auto-executes via Zerodha."
Q 1.3 · Required
What is your trading experience level?
⚠️ Beginner traders building fully automated systems without proven manual edge carry significant capital risk. We will flag this in your readiness score.
Q 1.4 · Required
What are the 3 most critical outcomes this system must deliver within 6 months?
Rank them by importance (1 = most critical). Be specific — use numbers where possible.
RankOutcomePriority
01
02
03
Q 1.5 · Optional
Do you have an existing system or product we should be aware of?
Existing tools / platforms in use
What must be preserved / migrated?
2.0
📊
Business Objectives & Success Metrics
Define measurable, time-bound success criteria. Without these, there is no way to evaluate whether the system is working. Vague goals = failed projects.
Q 2.1 · Required
What is your target ROI / profit expectation from this system?
Be realistic. Over-optimistic targets are the #1 red flag in trading system projects.
🔴 Red Flag Check: If your answer is "100%+ monthly returns" — this signals fundamental misunderstanding of algorithmic trading. Institutional quant funds target 15–40% annually. Any system promising more without verified multi-year backtest should be treated as suspect.
MetricTarget ValueTimeframe
Monthly return target (%)
Maximum acceptable drawdown
Win rate target
Sharpe Ratio target
Break-even timeline
Q 2.2 · Required
What is your total trading capital allocation for this system?
Q 2.3 · Required
How will you measure "system success" — select all that apply, then rank top 3:
Q 2.4 · Critical Check
Have you profitably traded this strategy MANUALLY before attempting to automate it?
This is the most important question in this entire document. Automating a losing strategy only makes you lose faster.
🔴 We strongly recommend a 3–6 month paper trading phase before deploying real capital, regardless of backtest results. Backtest ≠ live performance.
3.0
🧠
Trading Strategy & Decision Logic
This is the core engine of the entire system. Every architectural decision — data, latency, compute — cascades from your strategy definition. Be extremely precise.
STRATEGY DECISION FLOW
Market Data In ──▶ Signal Logic ──▶ [LONG / SHORT / HOLD] ──▶ Risk Filter ──▶ Execute / Skip
Each node below maps to a question. Define ALL of them before development begins.
Q 3.1 · Required
What trading strategy type(s) does this system implement? Select all that apply.
Q 3.2 · Required
Define your entry signal logic in detail.
List every condition that must be TRUE for the system to enter a trade. Use AND / OR logic explicitly. The more precise, the better the system.
Example (SMC Scalp on 15M chart): "Entry LONG when: (1) Price sweeps a liquidity level AND (2) breaks structure bullishly AND (3) RSI > 50 AND (4) EMA9 > EMA21 on the 1H chart AND (5) volume > 20-bar average."
Long (Buy) Entry Conditions
Short (Sell) Entry Conditions
Q 3.3 · Required
Define your exit strategy for each scenario:
Exit ScenarioLogic / RuleMethod
Take Profit (TP)
Stop Loss (SL)
Trailing Stop
Time-based Exit
Signal Reversal Exit
Q 3.4 · Required
What timeframes does your strategy use?
Select the PRIMARY execution timeframe and any CONFIRMATION timeframes used in multi-timeframe analysis (MTF).
EXECUTION TIMEFRAME (primary signal)
CONFIRMATION TIMEFRAMES (MTF bias)
Q 3.5 · Required
What indicators does your strategy depend on?
Custom indicator details (if any):
4.0
🌍
Market Scope & Instruments
Define the exact trading universe. Each market has unique data feeds, session hours, regulatory requirements, and liquidity profiles that directly affect system design.
Q 4.1 · Required
Which markets / exchanges will this system trade on?
Q 4.2 · Required
List the specific instruments / symbols this system must trade:
Be precise. Different symbols have different tick sizes, lot sizes, margin requirements, and liquidity.
SymbolExchangeAsset TypeApprox. Lot SizePriority
Q 4.3 · Required
What are the trading session constraints?
Active trading hours
Avoid trading during
Any blackout days / market holiday handling?
Q 4.4 · Optional
Will the system need to handle multiple instruments simultaneously?
ℹ️ 20+ simultaneous instruments requires significantly different architecture: parallel data pipelines, portfolio risk management, and distributed execution. This is a 3–5× complexity multiplier in development cost.
5.0
🗄️
Data Sources & Data Engineering
Garbage in, garbage out. Your strategy is only as good as the data feeding it. This section defines your entire data pipeline — quality, latency, cost, and reliability.
DATA PIPELINE ARCHITECTURE
Raw Source ──▶ Ingestion ──▶ Cleaning ──▶ Feature Eng. ──▶ Strategy Engine
↕ Storage ↕ Cache
Q 5.1 · Required
What type of market data does your strategy require?
Q 5.2 · Required
What data providers / APIs do you have access to or prefer?
ProviderData TypeStatusCost
Zerodha / Kite APIIndian equities, F&O, live + historical
Upstox APINSE/BSE equities, F&O
Alpha VantageGlobal stocks, Forex, Crypto (free/paid)
Yahoo Finance (yfinance)Global stocks, indices (free, delayed)
Polygon.ioUS market data, real-time
TrueData / DhanIndian market data, tick data
MetaTrader 5 (MT5)Forex, CFDs, commodities
Other
Q 5.3 · Required
What is your acceptable data latency for signal generation?
Latency directly determines infrastructure cost and architecture choice.
ℹ️ Sub-10ms latency requires co-location servers, FPGA, and specialized networking. Cost: $5,000–$50,000+/month. Most retail strategies work fine at 1–30 second latency at a fraction of the cost.
Q 5.4 · Required
How will missing or corrupt data be handled?
Q 5.5 · Required
How much historical data do you need stored and accessible?
Minimum historical depth required
Data granularity required
6.0
⏮️
Backtesting System Requirements
Backtesting is not optional — it is the scientific foundation of your strategy. Define exactly how rigorous your backtesting engine must be to produce trustworthy results.
BACKTESTING PIPELINE
Historical Data ──▶ Strategy Logic ──▶ Simulate Execution ──▶ Performance Report
↕ Slippage model     ↕ Commission calc     ↕ Walk-forward validation     ↕ Monte Carlo
Q 6.1 · Required
What backtesting engine or framework do you prefer?
Q 6.2 · Required
Which realism features must the backtester simulate?
Q 6.3 · Required
What validation methodology must be used?
Q 6.4 · Required
What performance metrics must the backtest report include?
Q 6.5 · Critical Check
What is your minimum acceptable backtest performance before going live?
MetricMinimum Threshold
Win Rate
Profit Factor
Max Drawdown
Total sample trades
Test period
7.0
⚡
Trade Execution & Automation
This section determines how and when trades are placed — from fully manual to 100% automated. Every sub-option has different risk, latency, and regulatory implications.
EXECUTION AUTOMATION SPECTRUM
Signal Generated ──▶ [Manual?] ──▶ User Reviews ──▶ User Places Order
│
↳ [Semi-auto?] ──▶ Alert + 1-click Execute
↳ [Full auto?] ──▶ API Executes Immediately
Q 7.1 · Required
What level of execution automation do you require?
⚠️ Fully automated execution carries regulatory risk in some markets (SEBI requires algo trading approval for certain frequencies). Confirm your broker permits API-based automated execution.
Q 7.2 · Required
Which broker(s) will the system execute orders through?
Q 7.3 · Required
What order types must the system support?
Q 7.4 · Required
How should the system handle execution failures?
Define the failsafe behavior when an order cannot be placed (API timeout, insufficient margin, market halted, etc.)
Failure ScenarioRequired System Response
Order rejected by broker
API timeout / connection lost
Insufficient margin
Market circuit breaker / halt
Partial fill received
Q 7.5 · Required
Position sizing — how should the system calculate trade size?
8.0
🛡️
Risk Management Framework
Risk management is non-negotiable. A system without a robust risk layer will eventually blow up your account — regardless of how good the signal is. Define every guard rail.
Q 8.1 · Required
Define your core risk parameters:
Risk LimitValue / RuleEnforcement
Max risk per single trade
Max daily loss (kill switch)
Max weekly loss
Max monthly drawdown
Max concurrent open positions
Max exposure per instrument
Max correlated positions
Q 8.2 · Required
Does the system need a global kill switch?
A kill switch immediately halts ALL trading and optionally closes all open positions when triggered.
Kill switch triggers
On kill switch activation:
Q 8.3 · Critical Check
🔴 Red Flag: What happens if the AI gives a BUY signal during a market crash?
This is a failure scenario question. The AI might not know about black swan events. How does your risk framework handle it?
🔴 If your answer is "rely on stop loss" only — understand that during a circuit breaker halt or flash crash, stop losses may not execute at your defined level (slippage). Plan for this explicitly.
9.0
🤖
AI / ML / LLM Integration
This is what separates a rule-based bot from an intelligent system. Define precisely what role AI plays — as the primary signal, as a filter, as an analyst, or as all three.
AI INTEGRATION MODES
Mode A: Rule-based indicators ──▶ AI validates / scores ──▶ Execute
Mode B: Raw market data ──▶ ML model predicts ──▶ Execute
Mode C: Indicators + news ──▶ LLM reasons ──▶ Human reviews ──▶ Execute
Q 9.1 · Required
What role should AI / ML play in this system?
Q 9.2 · Required
If using ML models, which types are preferred?
Model preference rationale:
Q 9.3 · Required
AI model training & retraining requirements:
RequirementYour Answer
Training data volume needed
Retraining frequency
Must the model explain its decisions?
Acceptable model latency
Where will model be hosted?
Q 9.4 · Critical Check
⚠️ How will you prevent AI from overfitting to backtested data?
Overfitting is the #1 killer of ML trading systems. A model trained on 5 years of data may have "learned" historical noise, not real patterns.
10.0
📈
Analysis, Insights & Reporting
Define what insights the system must surface, at what frequency, and in what format. Great systems don't just trade — they help you understand what's happening and why.
Q 10.1 · Required
What reports must the system generate automatically?
Q 10.2 · Required
What is the desired format for reports?
Q 10.3 · Optional
Should the AI generate natural language commentary on trades?
e.g., "Today's 3 losses were all caused by RSI divergence being ignored. Recommend adding RSI confirmation to entry logic."
11.0
🖥️
User Interface & Experience
Define how you interact with the system — the screens, controls, and information displays. Good UX separates a tool you use from a tool you trust.
Q 11.1 · Required
What interface(s) must the system provide?
Q 11.2 · Required
What must be visible on the main dashboard at all times?
Q 11.3 · Optional
Is there a reference UI or design style you want to match?
Examples: "Similar to Zerodha Kite — clean and minimal", "Bloomberg terminal aesthetic", "TradingView style dark charts", "Simple, like a mobile app"
12.0
🔌
Integrations & External Systems
A trading system doesn't exist in isolation. Map every external service it must connect to — for data, execution, communication, and compliance.
Q 12.1 · Required
Select all external integrations required:
IntegrationPurposeRequired?Have API keys?
Zerodha / Upstox / BrokerTrade execution + live data
Telegram Bot APIAlerts + commands
Claude / GPT APIAI analysis engine
TradingView WebhooksPine Script alert bridge
WhatsApp Business APITrade notifications
Google Sheets / NotionTrade journal / reporting
Email (SMTP)Reports + critical alerts
News API / RSSMarket news feed
13.0
⚡
Performance, Latency & Scalability
Define system performance benchmarks. A system that can't scale or slows down under load will fail at the worst possible moment — when the market is moving fast.
Q 13.1 · Required
System performance requirements:
MetricTarget SLAPriority
Signal generation latency
Order placement time
System uptime requirement
Max instruments monitored simultaneously
Backtest speed target
Dashboard load time
Q 13.2 · Optional
Future scalability requirements:
Expected growth in instruments (1 year)
Current 10
Expected trades per day (1 year)
Trades/day 20
14.0
🔒
Security, Compliance & Governance
A trading system holds API keys, capital, and sensitive financial data. Any breach is catastrophic. Define security and regulatory constraints before a single line of code is written.
Q 14.1 · Required
What security measures must be in place?
Q 14.2 · Required
Are there regulatory / compliance obligations?
⚠️ SEBI mandates that algo trading systems used for clients must be registered and audited. Personal-use systems have more flexibility but are still subject to monitoring. We recommend legal consultation before deploying any system that trades on behalf of others.
15.0
🔍
Monitoring, Logging & Alerting
Define how the system watches itself. A trading system running unmonitored is a liability. Every critical event needs a log, and every anomaly needs an alert.
Q 15.1 · Required
What events must trigger an immediate alert?
Q 15.2 · Required
Alert delivery channels — priority order:
ChannelUse forPriority
Telegram messageAll trade events
EmailDaily reports + critical alerts
WhatsAppTrade notifications
SMSCritical failures only
Push notification (mobile)All events
Dashboard alert bannerVisible on screen
16.0
☁️
Deployment, DevOps & Infrastructure
Where does this system live? Uptime, reliability, and cost are all determined by infrastructure choices. A system that runs on a laptop that sleeps will miss trades.
Q 16.1 · Required
Where will the system be hosted / deployed?
Q 16.2 · Required
DevOps and deployment requirements:
17.0
💰
Budget, Timeline & Constraints
Realistic budget and timeline expectations are the foundation of a healthy project. Misalignment here is the most common cause of failed deliveries.
Q 17.1 · Required
Development budget range:
Q 17.2 · Required
Timeline expectations:
MilestoneTarget Date / TimeframeFlexible?
MVP / Phase 1 (backtesting + signals)
Paper trading phase
Live deployment (Phase 2)
Full automation (Phase 3)
Q 17.3 · Required
Monthly operating cost budget (post-launch):
ℹ️ Typical monthly OpEx for a retail AI trading system: Server hosting $20–80, Data feeds $0–200, AI API calls $10–100, Monitoring tools $0–50. Total: $30–$430/month. Scale up for institutional needs.
18.0
⚠️
Edge Cases, Failures & Recovery
Murphy's Law applies aggressively in live trading. Define explicit responses to every foreseeable failure — because if it can go wrong, it will go wrong at the worst time.
Q 18.1 · Required
Failure scenario response plan:
ScenarioYour Required System Response
Internet goes down mid-trade
Broker API goes down
AI model gives extreme / impossible signal
Flash crash (10%+ drop in seconds)
Server / cloud hosting crashes
Strategy produces 10 consecutive losses
19.0
🚀
Future Roadmap & Expansion Vision
Architecture decisions made today affect what you can build tomorrow. Define your 1–3 year vision so the system is designed for extensibility from day one.
Q 19.1 · Optional
Which capabilities would you add in Phase 2 / 3?
Q 19.2 · Optional
Describe your 3-year vision for this system:
Example: "Phase 1: Personal AI trading bot for NSE equities. Phase 2: Add crypto + options. Phase 3: White-label as SaaS for other traders — 200 paying users at ₹2,000/month = ₹4L MRR."
20.0
💬
Open Discovery & Strategic Questions
Final open-ended questions that surface assumptions, unstated requirements, and strategic context that structured questions miss.
Q 20.1
What is the single biggest fear or concern you have about this project?
Q 20.2
What would make you immediately stop using this system?
Q 20.3
Is there anything about your trading or situation that is unusual that we should know?
Q 20.4
What assumptions have you made about this system that we should explicitly validate?
Examples of dangerous assumptions to challenge: "The AI will always make profitable signals", "Backtesting results will match live performance", "The broker API will always be available", "My strategy works in all market conditions"
SCORE
📊
System Readiness Scorecard
Your preparedness score across 8 domains. Complete more sections to improve your score. Click "Calculate Score" after finishing the questionnaire.
0
/ 100 READY
Complete the questionnaire to calculate your score
Strategy Clarity0–20 pts
Entry/exit logic defined
0/5
Timeframes specified
0/5
Indicators defined
0/5
Manual track record exists
0/5
Risk Management0–20 pts
Max loss limits defined
0/5
Position sizing method chosen
0/5
Kill switch designed
0/5
Failure scenarios answered
0/5
Data & Infrastructure0–15 pts
Data sources identified
0/5
Broker API access confirmed
0/5
Hosting plan chosen
0/5
Budget & Timeline0–15 pts
Realistic budget defined
0/5
Clear timeline milestones
0/5
ROI expectations realistic
0/5
OUTPUT
✅
Minimum Viable Input Checklist
These are the non-negotiable requirements before any development begins. Missing any of these = project cannot proceed safely.
✅ Must Have Before Dev Starts
🚀 Nice-to-Have (Phase 2+)
⚠️
Common Client Mistakes
The most frequent errors that derail AI trading system projects. Read these before proceeding.
#MistakeWhy It FailsPrevention
01 Automating before manual profitability Automation amplifies both profits AND losses. A losing strategy loses faster when automated. Require 6+ months of profitable manual trading before automation
02 Overfitting to backtesting Tuned to historical noise, not real patterns. Looks amazing in backtest, fails immediately live. Walk-forward testing + out-of-sample validation + paper trading
03 No kill switch or daily loss limit One bug = API fires 1000 orders. No limit = account blown in minutes. Hard-coded max daily loss + kill switch tested in staging
04 Unrealistic ROI expectations Expecting 50%/month → over-leveraging → one bad week = wipeout Target 2–6%/month consistently. Risk-adjusted returns beat raw returns.
05 Single point of failure (one broker, one server) Broker API goes down during high-volatility = can't exit positions = catastrophic loss Pre-placed bracket orders + fallback broker + heartbeat monitoring
06 Confusing signal quality with execution quality "AI gave a good signal but the system lost" — often a slippage or execution problem, not signal Log every signal vs. actual fill price. Measure slippage separately.
07 Deploying live without paper trading phase Bugs in execution logic only appear in live markets. Paper trading catches them safely. Mandatory 4–8 week paper trading on live market data before any real money
08 No version control on strategy code Changing a parameter and forgetting what worked before = can't reproduce past success Git for all code. Tag every backtest with the exact code version used.
When you're satisfied with your responses, click Export Responses to download a structured JSON summary, or Print PDF to generate a shareable document.
AI TRADING SYSTEM QUESTIONNAIRE · v2.0 · ALL RIGHTS RESERVED
Built for decision-ready requirement discovery. Not financial advice.