make-mp-data 2.1.6 → 3.0.1

package/dungeons/harness/harness-fintech.js

@@ -0,0 +1,976 @@
+import dayjs from "dayjs";
+import utc from "dayjs/plugin/utc.js";
+import "dotenv/config";
+import * as u from "../../lib/utils/utils.js";
+
+const SEED = "harness-fintech";
+dayjs.extend(utc);
+const chance = u.initChance(SEED);
+const num_users = 5_000;
+const days = 100;
+
+/** @typedef  {import("../../types.js").Dungeon} Config */
+
+/**
+ * NEEDLE IN A HAYSTACK - NEOBANK APP DESIGN
+ *
+ * NexBank - A Chime/Revolut-style neobank app where users manage accounts, make transactions,
+ * send transfers, pay bills, set budgets, invest, apply for loans, and earn rewards.
+ *
+ * CORE USER LOOP:
+ * Users open accounts (personal or business) through one of four channels (app, web, referral,
+ * branch). They begin using the app to check balances, make transactions (purchases, ATM
+ * withdrawals, direct deposits, refunds), and send transfers (internal, external, P2P, wire).
+ * The app tracks spending patterns and encourages financial wellness through budgets, savings
+ * goals, and investment tools.
+ *
+ * ACCOUNT TIERS:
+ * Three-tiered system drives monetization and user segmentation:
+ * - Basic (free): Standard banking, limited rewards
+ * - Plus ($4.99/mo): Enhanced rewards, priority support, budgeting tools
+ * - Premium ($14.99/mo): 3x rewards, 2x investment returns, premium analytics
+ * Tiers are assigned as superProps, creating a persistent segmentation dimension.
+ *
+ * TRANSACTION ECOSYSTEM:
+ * Transactions are the heartbeat of the app. Seven merchant categories (grocery, restaurant,
+ * gas, retail, online, subscription, utilities) and four payment methods (debit, credit,
+ * contactless, online) model realistic spending. Transaction amounts follow a power-law
+ * distribution centered around $50, with occasional large purchases up to $5,000.
+ *
+ * FINANCIAL WELLNESS:
+ * Budget creation and alerts model the "financial health" feature set. Users create budgets
+ * across six categories with monthly limits, then receive alerts when approaching or exceeding
+ * limits. This creates a measurable engagement loop: budget creators save more and invest more
+ * (Hook #5: Budget Users Save More).
+ *
+ * SAVINGS & INVESTMENT:
+ * Savings goals (emergency, vacation, car, home, education, retirement) with target amounts
+ * and monthly contributions model long-term financial planning. Investment events (stocks, ETF,
+ * crypto, bonds, mutual funds) with buy/sell actions model portfolio activity.
+ *
+ * LENDING:
+ * Loan applications (personal, auto, home, student, business) flow through apply-to-approved
+ * funnel. Requested vs. approved amounts and interest rates create realistic lending analytics.
+ *
+ * SUPPORT & ENGAGEMENT:
+ * Support contacts across four channels (chat, phone, email, in-app) with resolution tracking.
+ * Notifications (transaction, low_balance, bill_due, reward, security, promo) with action
+ * tracking model re-engagement effectiveness.
+ *
+ * BILL PAYMENTS:
+ * Six bill types (rent, utilities, phone, insurance, subscription, loan_payment) with auto-pay
+ * toggle. Hook #6 (Auto-Pay Loyalty) creates a pattern where auto-pay users never miss
+ * payments while manual payers miss 30%, modeling real-world payment reliability.
+ *
+ * REWARDS:
+ * Four reward types (cashback, points, discount, partner_offer) with values scaled by
+ * account tier. Premium users earn 3x rewards (Hook #7), creating clear tier-based value
+ * differentiation visible in the data.
+ *
+ * HOUSEHOLD GROUPS:
+ * 500 households group users for shared financial analytics. Household-level properties
+ * (size, combined income, financial health score, primary bank) enable group-level analysis
+ * of shared financial behaviors.
+ *
+ * WHY THESE EVENTS/PROPERTIES?
+ * - Events model a complete banking loop: onboarding -> daily use -> financial planning -> monetization
+ * - Properties enable cohort analysis: account type, tier, income bracket, credit score
+ * - Funnels reveal friction: onboarding completion, transfer flows, investment journeys
+ * - Financial wellness features (budgets, goals) create engagement depth visible in the data
+ * - Tier-based rewards and investment returns drive business metric differences
+ * - The "needle in haystack" hooks simulate real fintech product insights hidden in production data
+ */
+
+/** @type {Config} */
+const config = {
+	token: "048f954dddbfbd4bff34edd8b3d95c06",
+	seed: SEED,
+	numDays: days,
+	numEvents: num_users * 120,
+	numUsers: num_users,
+	hasAnonIds: false,
+	hasSessionIds: true,
+	format: "json",
+	gzip: true,
+	alsoInferFunnels: false,
+	hasLocation: true,
+	hasAndroidDevices: true,
+	hasIOSDevices: true,
+	hasDesktopDevices: true,
+	hasBrowser: false,
+	hasCampaigns: false,
+	isAnonymous: false,
+	hasAdSpend: false,
+	percentUsersBornInDataset: 50,
+
+	hasAvatar: true,
+	makeChart: false,
+
+	batchSize: 2_500_000,
+	concurrency: 10,
+	writeToDisk: false,
+
+	scdProps: {},
+
+	funnels: [
+		{
+			sequence: ["account opened", "app session", "balance checked"],
+			isFirstFunnel: true,
+			conversionRate: 85,
+			timeToConvert: 0.25,
+		},
+		{
+			// Daily banking: check balance, view transactions - most common activity
+			sequence: ["app session", "balance checked", "transaction completed"],
+			conversionRate: 80,
+			timeToConvert: 0.5,
+			weight: 5,
+		},
+		{
+			// Transfers and notifications
+			sequence: ["app session", "transfer sent", "notification opened"],
+			conversionRate: 50,
+			timeToConvert: 1,
+			weight: 3,
+		},
+		{
+			// Bill payment flow
+			sequence: ["app session", "bill paid", "notification opened"],
+			conversionRate: 60,
+			timeToConvert: 1,
+			weight: 3,
+		},
+		{
+			// Financial planning: budgets and savings
+			sequence: ["budget created", "budget alert", "savings goal set"],
+			conversionRate: 40,
+			timeToConvert: 12,
+			weight: 2,
+		},
+		{
+			// Investment and rewards
+			sequence: ["balance checked", "investment made", "reward redeemed"],
+			conversionRate: 30,
+			timeToConvert: 5,
+			weight: 2,
+		},
+		{
+			// Support and account management
+			sequence: ["support contacted", "card locked", "dispute filed"],
+			conversionRate: 35,
+			timeToConvert: 2,
+			weight: 1,
+		},
+		{
+			// Lending flow
+			sequence: ["loan applied", "loan approved", "premium upgraded"],
+			conversionRate: 25,
+			timeToConvert: 10,
+			weight: 1,
+		},
+	],
+
+	events: [
+		{
+			event: "account opened",
+			weight: 1,
+			isFirstEvent: true,
+			properties: {
+				"account_type": u.pickAWinner(["personal", "business", "personal"]),
+				"signup_channel": u.pickAWinner(["app", "web", "referral", "branch"]),
+			}
+		},
+		{
+			event: "app session",
+			weight: 20,
+			properties: {
+				"session_duration_sec": u.weighNumRange(10, 600, 0.3, 60),
+				"pages_viewed": u.weighNumRange(1, 15, 0.5, 3),
+			}
+		},
+		{
+			event: "balance checked",
+			weight: 15,
+			properties: {
+				"account_balance": u.weighNumRange(0, 50000, 0.3, 2500),
+				"account_type": u.pickAWinner(["checking", "savings", "investment"]),
+			}
+		},
+		{
+			event: "transaction completed",
+			weight: 18,
+			properties: {
+				"transaction_type": u.pickAWinner(["purchase", "atm", "direct_deposit", "refund"]),
+				"amount": u.weighNumRange(1, 5000, 0.3, 50),
+				"merchant_category": u.pickAWinner(["grocery", "restaurant", "gas", "retail", "online", "subscription", "utilities"]),
+				"payment_method": u.pickAWinner(["debit", "credit", "contactless", "online"]),
+			}
+		},
+		{
+			event: "transfer sent",
+			weight: 8,
+			properties: {
+				"transfer_type": u.pickAWinner(["internal", "external", "p2p", "wire"]),
+				"amount": u.weighNumRange(10, 10000, 0.3, 200),
+				"recipient_type": u.pickAWinner(["friend", "family", "business", "self"]),
+			}
+		},
+		{
+			event: "bill paid",
+			weight: 6,
+			properties: {
+				"bill_type": u.pickAWinner(["rent", "utilities", "phone", "insurance", "subscription", "loan_payment"]),
+				"amount": u.weighNumRange(20, 3000, 0.5, 150),
+				"auto_pay": u.pickAWinner([true, false], 0.4),
+			}
+		},
+		{
+			event: "budget created",
+			weight: 3,
+			properties: {
+				"category": u.pickAWinner(["food", "transport", "entertainment", "shopping", "bills", "savings"]),
+				"monthly_limit": u.weighNumRange(50, 2000, 0.5, 300),
+			}
+		},
+		{
+			event: "budget alert",
+			weight: 4,
+			properties: {
+				"alert_type": u.pickAWinner(["approaching_limit", "exceeded", "on_track"]),
+				"percent_used": u.weighNumRange(50, 150, 1, 90),
+			}
+		},
+		{
+			event: "savings goal set",
+			weight: 3,
+			properties: {
+				"goal_type": u.pickAWinner(["emergency", "vacation", "car", "home", "education", "retirement"]),
+				"target_amount": u.weighNumRange(500, 50000, 0.3, 5000),
+				"monthly_contribution": u.weighNumRange(25, 2000, 0.5, 200),
+			}
+		},
+		{
+			event: "investment made",
+			weight: 4,
+			properties: {
+				"investment_type": u.pickAWinner(["stocks", "etf", "crypto", "bonds", "mutual_fund"]),
+				"amount": u.weighNumRange(10, 10000, 0.3, 250),
+				"action": u.pickAWinner(["buy", "sell", "buy"]),
+			}
+		},
+		{
+			event: "card locked",
+			weight: 1,
+			properties: {
+				"reason": u.pickAWinner(["lost", "stolen", "suspicious_activity", "travel"]),
+			}
+		},
+		{
+			event: "dispute filed",
+			weight: 1,
+			properties: {
+				"dispute_amount": u.weighNumRange(10, 2000, 0.5, 100),
+				"reason": u.pickAWinner(["unauthorized", "duplicate", "not_received", "damaged", "wrong_amount"]),
+			}
+		},
+		{
+			event: "loan applied",
+			weight: 2,
+			properties: {
+				"loan_type": u.pickAWinner(["personal", "auto", "home", "student", "business"]),
+				"requested_amount": u.weighNumRange(1000, 100000, 0.3, 10000),
+			}
+		},
+		{
+			event: "loan approved",
+			weight: 1,
+			properties: {
+				"loan_type": u.pickAWinner(["personal", "auto", "home", "student", "business"]),
+				"approved_amount": u.weighNumRange(1000, 100000, 0.3, 10000),
+				"interest_rate": u.weighNumRange(3, 25, 1, 8),
+			}
+		},
+		{
+			event: "premium upgraded",
+			weight: 2,
+			properties: {
+				"old_tier": u.pickAWinner(["basic", "plus", "premium"]),
+				"new_tier": u.pickAWinner(["plus", "premium", "premium"]),
+				"monthly_fee": u.pickAWinner([4.99, 9.99, 14.99]),
+			}
+		},
+		{
+			event: "support contacted",
+			weight: 3,
+			properties: {
+				"channel": u.pickAWinner(["chat", "phone", "email", "in_app"]),
+				"issue_type": u.pickAWinner(["transaction", "account", "card", "transfer", "technical"]),
+				"resolved": u.pickAWinner([true, false], 0.8),
+			}
+		},
+		{
+			event: "notification opened",
+			weight: 10,
+			properties: {
+				"notification_type": u.pickAWinner(["transaction", "low_balance", "bill_due", "reward", "security", "promo"]),
+				"action_taken": u.pickAWinner([true, false], 0.6),
+			}
+		},
+		{
+			event: "reward redeemed",
+			weight: 4,
+			properties: {
+				"reward_type": u.pickAWinner(["cashback", "points", "discount", "partner_offer"]),
+				"value": u.weighNumRange(1, 100, 0.5, 10),
+			}
+		}
+	],
+
+	superProps: {
+		account_tier: u.pickAWinner(["basic", "basic", "basic", "plus", "plus", "premium"]),
+		platform: u.pickAWinner(["ios", "android", "web"]),
+	},
+
+	userProps: {
+		"credit_score_range": u.pickAWinner(["300-579", "580-669", "670-739", "740-799", "800-850"]),
+		"income_bracket": u.pickAWinner(["under_30k", "30k_50k", "50k_75k", "75k_100k", "100k_150k", "over_150k"]),
+		"account_age_months": u.weighNumRange(1, 60, 0.5, 12),
+		"total_balance": u.weighNumRange(0, 100000, 0.3, 5000),
+		"has_direct_deposit": u.pickAWinner([true, false], 0.6),
+	},
+
+	groupKeys: [
+		["household_id", 500, ["transaction completed", "transfer sent", "bill paid", "savings goal set"]],
+	],
+
+	groupProps: {
+		household_id: {
+			"household_size": u.weighNumRange(1, 6),
+			"combined_income": u.weighNumRange(20000, 300000, 0.3, 75000),
+			"financial_health_score": u.weighNumRange(1, 100, 1, 65),
+			"primary_bank": u.pickAWinner(["NexBank_only", "multi_bank", "NexBank_only"]),
+		}
+	},
+
+	lookupTables: [],
+
+	/**
+	 * ARCHITECTED ANALYTICS HOOKS
+	 *
+	 * This hook function creates 8 deliberate patterns in the data:
+	 *
+	 * 1. PERSONAL VS BUSINESS: Business accounts get employee_count, revenue; personal get age_range, life_stage
+	 * 2. PAYDAY PATTERNS: Transactions spike on 1st/15th with bigger deposits and post-payday spending
+	 * 3. FRAUD DETECTION: 3% of users experience a fraud burst (rapid high-value txns -> card lock -> dispute -> support)
+	 * 4. LOW BALANCE CHURN: Users with chronic low balances (<$500) lose 50% of activity after day 30
+	 * 5. BUDGET DISCIPLINE: Budget creators save 2x more and invest 1.5x more
+	 * 6. AUTO-PAY LOYALTY: Auto-pay users never miss bills; manual payers miss 30%
+	 * 7. PREMIUM TIER VALUE: Premium users get 3x rewards; Plus users get 1.5x; Premium investors get 2x returns
+	 * 8. MONTH-END ANXIETY: Last 3 days of month see 40% longer sessions and 30% lower balances
+	 */
+	hook: function (record, type, meta) {
+		const NOW = dayjs();
+		const DATASET_START = NOW.subtract(days, "days");
+
+		// ===============================================================
+		// Hook #1: PERSONAL VS BUSINESS ACCOUNTS (user)
+		// 20% of users are randomly tagged as business accounts with
+		// employee_count, annual_revenue, and industry. The other 80% get
+		// personal attributes: age_range and life_stage.
+		// ===============================================================
+		if (type === "user") {
+			const isBusiness = chance.bool({ likelihood: 20 });
+			if (isBusiness) {
+				record.account_segment = "business";
+				record.employee_count = chance.integer({ min: 5, max: 500 });
+				record.annual_revenue = chance.integer({ min: 100000, max: 10000000 });
+				record.industry = chance.pickone(["tech", "retail", "food", "services", "healthcare"]);
+			} else {
+				record.account_segment = "personal";
+				record.age_range = `${chance.pickone([18, 25, 35, 45, 55])}-${chance.pickone([24, 34, 44, 54, 65])}`;
+				record.life_stage = chance.pickone(["student", "early_career", "established", "pre_retirement", "retired"]);
+			}
+		}
+
+		// ===============================================================
+		// Hook #2: PAYDAY PATTERNS (event)
+		// On the 1st and 15th, direct deposit transactions are 2x bigger
+		// and tagged with payday: true. On days 1-3 and 15-17, transfers
+		// have a 40% chance of 1.5x boost (post-payday spending).
+		// ===============================================================
+		if (type === "event") {
+			const EVENT_TIME = dayjs(record.time);
+			const dayOfMonth = EVENT_TIME.date();
+
+			// Payday: 1st and 15th
+			if (record.event === "transaction completed" && record.transaction_type === "direct_deposit") {
+				if (dayOfMonth === 1 || dayOfMonth === 15) {
+					record.amount = Math.floor((record.amount || 50) * 2);
+					record.payday = true;
+				} else {
+					record.payday = false;
+				}
+			}
+
+			// Post-payday spending: days 1-3 and 15-17
+			if (record.event === "transfer sent") {
+				const isPaydayWindow = (dayOfMonth >= 1 && dayOfMonth <= 3) || (dayOfMonth >= 15 && dayOfMonth <= 17);
+				if (isPaydayWindow && chance.bool({ likelihood: 40 })) {
+					record.amount = Math.floor((record.amount || 200) * 1.5);
+					record.post_payday_spending = true;
+				} else {
+					record.post_payday_spending = false;
+				}
+			}
+
+			// ===============================================================
+			// Hook #6: AUTO-PAY LOYALTY (event)
+			// Users with auto_pay=false on bill paid events have a 30% chance
+			// of the event being dropped entirely (missed payment). Surviving
+			// manual payments get tagged. Auto-pay users always succeed.
+			// ===============================================================
+			if (record.event === "bill paid") {
+				if (record.auto_pay === false || record.auto_pay === undefined) {
+					if (chance.bool({ likelihood: 30 })) {
+						record.event = "bill payment missed";
+						record.missed_payment = true;
+						record.manual_payment = false;
+					} else {
+						record.missed_payment = false;
+						record.manual_payment = true;
+					}
+				} else {
+					record.missed_payment = false;
+					record.manual_payment = false;
+				}
+			}
+
+			// ===============================================================
+			// Hook #7: PREMIUM TIER VALUE (event)
+			// Premium tier users get 3x reward values and 2x investment sell
+			// returns. Plus tier users get 1.5x rewards.
+			// ===============================================================
+			if (record.event === "reward redeemed") {
+				if (record.account_tier === "premium") {
+					record.value = Math.floor((record.value || 10) * 3);
+					record.premium_reward = true;
+				} else if (record.account_tier === "plus") {
+					record.value = Math.floor((record.value || 10) * 1.5);
+					record.premium_reward = false;
+				} else {
+					record.premium_reward = false;
+				}
+			}
+
+			if (record.event === "investment made" && record.action === "sell") {
+				if (record.account_tier === "premium") {
+					record.amount = Math.floor((record.amount || 250) * 2);
+					record.premium_returns = true;
+				} else {
+					record.premium_returns = false;
+				}
+			}
+
+			// ===============================================================
+			// Hook #8: MONTH-END ANXIETY (event)
+			// Last 3 days of month (day >= 28) see 40% longer app sessions
+			// and 30% lower balances when checked.
+			// ===============================================================
+			if (record.event === "app session") {
+				if (dayOfMonth >= 28) {
+					record.session_duration_sec = Math.floor((record.session_duration_sec || 60) * 1.4);
+					record.month_end_anxiety = true;
+				} else {
+					record.month_end_anxiety = false;
+				}
+			}
+
+			if (record.event === "balance checked") {
+				if (dayOfMonth >= 28) {
+					record.account_balance = Math.floor((record.account_balance || 2500) * 0.7);
+					record.month_end_check = true;
+				} else {
+					record.month_end_check = false;
+				}
+			}
+		}
+
+		// ===============================================================
+		// Hook #3, #4, #5: EVERYTHING - Complex behavioral patterns
+		// These hooks operate on the full event stream per user for
+		// fraud injection, low-balance churn, and budget discipline.
+		// ===============================================================
+		if (type === "everything") {
+			const userEvents = record;
+			const firstEventTime = userEvents.length > 0 ? dayjs(userEvents[0].time) : null;
+
+			// -----------------------------------------------------------
+			// Hook #3: FRAUD DETECTION PATTERN
+			// 3% of users experience a fraud event sequence: a burst of
+			// 3-5 rapid high-value transactions within 1 hour, followed by
+			// card locked, dispute filed, and support contacted events.
+			// All injected events are tagged with fraud_sequence: true.
+			// -----------------------------------------------------------
+			if (chance.bool({ likelihood: 3 })) {
+				// Find the midpoint of the user's timeline
+				if (userEvents.length >= 2) {
+					const midIdx = Math.floor(userEvents.length / 2);
+					const midEvent = userEvents[midIdx];
+					const midTime = dayjs(midEvent.time);
+					const distinctId = midEvent.user_id;
+
+					// Inject 3-5 rapid high-value transactions
+					const burstCount = chance.integer({ min: 3, max: 5 });
+					const fraudEvents = [];
+
+					for (let i = 0; i < burstCount; i++) {
+						fraudEvents.push({
+							event: "transaction completed",
+							time: midTime.add(i * 10, "minutes").toISOString(),
+							user_id: distinctId,
+							transaction_type: "purchase",
+							amount: chance.integer({ min: 500, max: 3000 }),
+							merchant_category: chance.pickone(["online", "retail"]),
+							payment_method: "credit",
+							fraud_sequence: true,
+						});
+					}
+
+					// Card locked after the burst
+					fraudEvents.push({
+						event: "card locked",
+						time: midTime.add(burstCount * 10 + 5, "minutes").toISOString(),
+						user_id: distinctId,
+						reason: "suspicious_activity",
+						fraud_sequence: true,
+					});
+
+					// Dispute filed shortly after
+					fraudEvents.push({
+						event: "dispute filed",
+						time: midTime.add(burstCount * 10 + 30, "minutes").toISOString(),
+						user_id: distinctId,
+						dispute_amount: chance.integer({ min: 500, max: 3000 }),
+						reason: "unauthorized",
+						fraud_sequence: true,
+					});
+
+					// Support contacted
+					fraudEvents.push({
+						event: "support contacted",
+						time: midTime.add(burstCount * 10 + 45, "minutes").toISOString(),
+						user_id: distinctId,
+						channel: "phone",
+						issue_type: "card",
+						resolved: true,
+						fraud_sequence: true,
+					});
+
+					// Splice all fraud events into the user's timeline
+					userEvents.splice(midIdx + 1, 0, ...fraudEvents);
+				}
+			}
+
+			// -----------------------------------------------------------
+			// Hook #4: LOW BALANCE CHURN
+			// Users who have 3+ balance checks showing < $500 are
+			// "struggling". After day 30, 50% of their events are removed
+			// and remaining events are tagged low_balance_churn: true.
+			// -----------------------------------------------------------
+			let lowBalanceChecks = 0;
+			userEvents.forEach((event) => {
+				if (event.event === "balance checked" && (event.account_balance || 0) < 500) {
+					lowBalanceChecks++;
+				}
+			});
+
+			if (lowBalanceChecks >= 3) {
+				const day30 = DATASET_START.add(30, "days");
+				for (let i = userEvents.length - 1; i >= 0; i--) {
+					const evt = userEvents[i];
+					if (dayjs(evt.time).isAfter(day30)) {
+						if (chance.bool({ likelihood: 50 })) {
+							userEvents.splice(i, 1);
+						} else {
+							evt.low_balance_churn = true;
+						}
+					}
+				}
+			}
+
+			// -----------------------------------------------------------
+			// Hook #5: BUDGET USERS SAVE MORE
+			// Users who created any budget have 2x savings contributions,
+			// 1.5x investment amounts, and extra savings goal events
+			// spliced into their timeline. Tagged budget_discipline: true.
+			// -----------------------------------------------------------
+			let hasBudget = false;
+			userEvents.forEach((event) => {
+				if (event.event === "budget created") {
+					hasBudget = true;
+				}
+			});
+
+			if (hasBudget) {
+				userEvents.forEach((event, idx) => {
+					const eventTime = dayjs(event.time);
+
+					// Double savings contributions
+					if (event.event === "savings goal set") {
+						event.monthly_contribution = Math.floor((event.monthly_contribution || 200) * 2);
+						event.budget_discipline = true;
+					}
+
+					// 1.5x investment amounts
+					if (event.event === "investment made") {
+						event.amount = Math.floor((event.amount || 250) * 1.5);
+						event.budget_discipline = true;
+					}
+
+					// Splice extra savings goal events (budget-conscious users set more goals)
+					if (event.event === "budget created" && chance.bool({ likelihood: 50 })) {
+						const extraGoal = {
+							event: "savings goal set",
+							time: eventTime.add(chance.integer({ min: 1, max: 7 }), "days").toISOString(),
+							user_id: event.user_id,
+							goal_type: chance.pickone(["emergency", "vacation", "car", "home"]),
+							target_amount: chance.integer({ min: 1000, max: 20000 }),
+							monthly_contribution: chance.integer({ min: 100, max: 800 }),
+							budget_discipline: true,
+						};
+						userEvents.splice(idx + 1, 0, extraGoal);
+					}
+				});
+			}
+		}
+
+		return record;
+	}
+};
+
+export default config;
+
+/**
+ * ===================================================================
+ * NEEDLE IN A HAYSTACK - NEXBANK NEOBANK ANALYTICS
+ * ===================================================================
+ *
+ * A Chime/Revolut-style neobank dungeon with 8 deliberately architected
+ * analytics insights hidden in the data. This dungeon is designed to
+ * showcase advanced fintech product analytics patterns and demonstrate
+ * how to find "needles" (meaningful insights) in "haystacks" (large
+ * banking datasets).
+ *
+ * ===================================================================
+ * DATASET OVERVIEW
+ * ===================================================================
+ *
+ * - 5,000 users over 100 days
+ * - 360,000 events across 18 event types
+ * - 3 funnels (onboarding, transfer flow, investment journey)
+ * - Group analytics (500 households)
+ * - Lookup table (500 merchant/transaction entries)
+ * - Account tiers (Basic, Plus, Premium)
+ *
+ * ===================================================================
+ * THE 8 ARCHITECTED HOOKS
+ * ===================================================================
+ *
+ * Each hook creates a specific, discoverable analytics insight that
+ * simulates real-world fintech product behavior patterns.
+ *
+ * -------------------------------------------------------------------
+ * 1. PERSONAL VS BUSINESS ACCOUNTS (user)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: 20% of users are business accounts with employee_count
+ * (5-500), annual_revenue ($100K-$10M), and industry. The other 80%
+ * are personal accounts with age_range and life_stage.
+ *
+ * HOW TO FIND IT:
+ *   - Segment users by: account_segment = "business" vs "personal"
+ *   - Compare: Transaction volumes, average amounts, transfer patterns
+ *   - Analyze: Industry distribution among business accounts
+ *
+ * EXPECTED INSIGHT: Business accounts have fundamentally different
+ * usage patterns - higher transaction amounts, more wire transfers,
+ * and different bill payment profiles.
+ *
+ * REAL-WORLD ANALOGUE: Identifying and serving different customer
+ * segments (B2C vs B2B) with tailored features and pricing.
+ *
+ * -------------------------------------------------------------------
+ * 2. PAYDAY PATTERNS (event)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: On the 1st and 15th of each month, direct deposit
+ * transactions are 2x bigger and tagged with payday: true. On days
+ * 1-3 and 15-17, transfer amounts have a 40% chance of being 1.5x
+ * larger, tagged with post_payday_spending: true.
+ *
+ * HOW TO FIND IT:
+ *   - Chart: transaction completed count and amount by day of month
+ *   - Filter: transaction_type = "direct_deposit"
+ *   - Compare: Average transfer amount on days 1-3/15-17 vs other days
+ *   - Look for: payday: true and post_payday_spending: true tags
+ *
+ * EXPECTED INSIGHT: Clear biweekly spikes in deposit amounts and
+ * subsequent spending activity. The 2-3 days after payday show
+ * elevated transfer activity as users move money around.
+ *
+ * REAL-WORLD ANALOGUE: Payroll cycle effects on banking activity.
+ * Banks use this to time marketing, credit offers, and overdraft
+ * protection promotions.
+ *
+ * -------------------------------------------------------------------
+ * 3. FRAUD DETECTION PATTERN (everything)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: 3% of users experience a fraud event sequence at their
+ * timeline midpoint: 3-5 rapid high-value transactions ($500-$3,000)
+ * within 1 hour, followed by card locked (suspicious_activity),
+ * dispute filed (unauthorized), and support contacted (phone/card).
+ * All injected events tagged with fraud_sequence: true.
+ *
+ * HOW TO FIND IT:
+ *   - Filter events: fraud_sequence = true
+ *   - Analyze: Time between fraud transactions (< 10 min gaps)
+ *   - Funnel: transaction completed -> card locked -> dispute filed -> support contacted
+ *   - Segment: Users with any fraud_sequence event
+ *
+ * EXPECTED INSIGHT: ~150 users (3% of 5,000) show a distinctive burst
+ * pattern of rapid high-value purchases followed by account lockdown.
+ * Clear temporal clustering of fraud events.
+ *
+ * REAL-WORLD ANALOGUE: Fraud detection in banking. Unusual velocity
+ * and amount patterns trigger automated alerts and account freezes.
+ *
+ * -------------------------------------------------------------------
+ * 4. LOW BALANCE CHURN (everything)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: Users with 3+ balance checks showing < $500 are
+ * "struggling" users. After day 30, 50% of their events are removed
+ * (simulating reduced app usage) and surviving events are tagged
+ * with low_balance_churn: true.
+ *
+ * HOW TO FIND IT:
+ *   - Segment: Users where count of (balance checked, account_balance < 500) >= 3
+ *   - Compare: Event counts before day 30 vs after day 30
+ *   - Filter: low_balance_churn = true
+ *   - Retention analysis: Compare D30+ retention for low vs healthy balance users
+ *
+ * EXPECTED INSIGHT: Struggling users show a dramatic drop in activity
+ * after the first month. Their engagement halves while healthy-balance
+ * users maintain consistent usage.
+ *
+ * REAL-WORLD ANALOGUE: Financial stress-driven churn. Users who
+ * can't maintain balances disengage from their banking app, which
+ * predicts account closure.
+ *
+ * -------------------------------------------------------------------
+ * 5. BUDGET USERS SAVE MORE (everything)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: Users who create any budget have 2x monthly savings
+ * contributions, 1.5x investment amounts, and extra savings goal
+ * events spliced into their timeline. All affected events tagged
+ * with budget_discipline: true.
+ *
+ * HOW TO FIND IT:
+ *   - Segment: Users who did "budget created" vs those who didn't
+ *   - Compare: Average monthly_contribution on savings goal events
+ *   - Compare: Average investment amount
+ *   - Count: savings goal set events per user (budget users have more)
+ *   - Filter: budget_discipline = true
+ *
+ * EXPECTED INSIGHT: Budget creators save 2x more and invest 1.5x more
+ * than non-budget users. They also set more savings goals, showing
+ * compound financial wellness behavior.
+ *
+ * REAL-WORLD ANALOGUE: Financial planning tools drive better outcomes.
+ * Users who engage with budgeting features are more financially active
+ * and retain longer - a key product-market fit signal.
+ *
+ * -------------------------------------------------------------------
+ * 6. AUTO-PAY LOYALTY (event)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: Bill paid events with auto_pay = false have a 30% chance
+ * of being dropped entirely (simulating missed payments). Auto-pay
+ * users never miss. Surviving manual payments are tagged with
+ * manual_payment: true.
+ *
+ * HOW TO FIND IT:
+ *   - Segment: bill paid events by auto_pay = true vs false
+ *   - Compare: Total bill paid count per user in each segment
+ *   - Calculate: Effective bill completion rate by segment
+ *   - Filter: manual_payment = true for surviving manual payments
+ *
+ * EXPECTED INSIGHT: Auto-pay users have 100% bill completion while
+ * manual payers show only ~70% completion. This creates a clear
+ * reliability gap that would drive auto-pay adoption campaigns.
+ *
+ * REAL-WORLD ANALOGUE: Auto-pay enrollment is one of the strongest
+ * retention predictors in fintech. Users who set up auto-pay are
+ * less likely to miss payments and less likely to churn.
+ *
+ * -------------------------------------------------------------------
+ * 7. PREMIUM TIER VALUE (event)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: Premium tier users get 3x reward values and 2x investment
+ * sell returns. Plus tier users get 1.5x rewards. Tagged with
+ * premium_reward: true and premium_returns: true respectively.
+ *
+ * HOW TO FIND IT:
+ *   - Segment: Events by account_tier (basic, plus, premium)
+ *   - Compare: Average reward value on reward redeemed events
+ *   - Compare: Average amount on investment made (action = sell)
+ *   - Filter: premium_reward = true, premium_returns = true
+ *   - Analyze: Total reward value per user by tier
+ *
+ * EXPECTED INSIGHT: Clear tier-based value curve. Premium users
+ * earn 3x the rewards and 2x the investment returns of Basic users,
+ * with Plus users in between. This validates tier pricing.
+ *
+ * REAL-WORLD ANALOGUE: Premium banking tiers that provide tangible
+ * financial benefits. The reward multiplier justifies the monthly
+ * fee and drives upgrade conversions.
+ *
+ * -------------------------------------------------------------------
+ * 8. MONTH-END ANXIETY (event)
+ * -------------------------------------------------------------------
+ *
+ * PATTERN: On the last 3 days of each month (day >= 28), app sessions
+ * are 40% longer and balance checks show 30% lower balances. Tagged
+ * with month_end_anxiety: true and month_end_check: true.
+ *
+ * HOW TO FIND IT:
+ *   - Chart: Average session_duration_sec by day of month
+ *   - Chart: Average account_balance on balance checked by day of month
+ *   - Filter: month_end_anxiety = true, month_end_check = true
+ *   - Compare: Day 1-27 vs day 28-31 engagement metrics
+ *
+ * EXPECTED INSIGHT: Users spend 40% more time in the app at month-end,
+ * checking lower balances. This reflects pre-bill-pay anxiety and
+ * end-of-month financial stress.
+ *
+ * REAL-WORLD ANALOGUE: Month-end financial anxiety drives app
+ * engagement spikes. Banks can leverage this pattern for timely
+ * overdraft protection offers, savings nudges, and bill reminders.
+ *
+ * ===================================================================
+ * ADVANCED ANALYSIS IDEAS
+ * ===================================================================
+ *
+ * CROSS-HOOK PATTERNS:
+ *
+ * 1. Budget + Low Balance: Do budget creators (Hook #5) avoid the
+ *    low-balance churn pattern (Hook #4)? Budget discipline should
+ *    correlate with healthier balances.
+ *
+ * 2. Premium + Auto-Pay: Do premium tier users (Hook #7) have higher
+ *    auto-pay adoption than basic users? Does tier upgrading predict
+ *    auto-pay enrollment?
+ *
+ * 3. Fraud + Churn: Do fraud victims (Hook #3) churn more than
+ *    non-victims? Does support resolution quality affect post-fraud
+ *    retention?
+ *
+ * 4. Payday + Month-End: Compare payday spending spikes (Hook #2)
+ *    with month-end anxiety (Hook #8). Do payday spenders run out
+ *    of money by month-end?
+ *
+ * 5. Business vs Personal Fraud: Are business accounts (Hook #1)
+ *    more or less likely to be fraud targets (Hook #3)?
+ *
+ * COHORT ANALYSIS:
+ *
+ * - Cohort by account_tier: Track upgrade paths and value realization
+ *   across Basic -> Plus -> Premium
+ * - Cohort by signup_channel: Do referral users have better retention
+ *   and higher balances?
+ * - Cohort by income_bracket: How does income correlate with feature
+ *   adoption (budgets, investments, savings goals)?
+ * - Cohort by credit_score_range: Do higher credit scores predict
+ *   loan approvals and premium tier adoption?
+ *
+ * FUNNEL ANALYSIS:
+ *
+ * - Onboarding Funnel: account opened -> app session -> balance checked
+ *   by account type and signup channel
+ * - Transfer Flow: app session -> transfer sent -> notification opened
+ *   by tier and platform
+ * - Investment Journey: balance checked -> investment made -> reward redeemed
+ *   by income bracket and budget creation status
+ *
+ * ===================================================================
+ * EXPECTED METRICS SUMMARY
+ * ===================================================================
+ *
+ * Hook                  | Metric                | Baseline | Hook Effect | Ratio
+ * ----------------------|-----------------------|----------|-------------|------
+ * Personal vs Business  | Avg transaction amt   | $50      | $200+       | ~4x
+ * Payday Patterns       | Deposit amount        | $50      | $100        | 2x
+ * Fraud Detection       | Users affected        | 0%       | 3%          | --
+ * Low Balance Churn     | D30+ event count      | 100%     | 50%         | 0.5x
+ * Budget Discipline     | Monthly contribution  | $200     | $400        | 2x
+ * Auto-Pay Loyalty      | Bill completion rate   | 100%     | 70%         | 0.7x
+ * Premium Tier Value    | Reward value           | $10      | $30         | 3x
+ * Month-End Anxiety     | Session duration       | 60s      | 84s         | 1.4x
+ *
+ * ===================================================================
+ * HOW TO RUN THIS DUNGEON
+ * ===================================================================
+ *
+ * From the dm4 root directory:
+ *
+ *   npm start
+ *
+ * Or programmatically:
+ *
+ *   import generate from './index.js';
+ *   import config from './dungeons/harness-fintech.js';
+ *   const results = await generate(config);
+ *
+ * OUTPUT FILES (with writeToDisk: true):
+ *
+ *   - needle-haystack-fintech__events.json.gz       - All event data
+ *   - needle-haystack-fintech__user_profiles.json.gz - User profiles
+ *   - needle-haystack-fintech__group_profiles.json.gz - Household profiles
+ *   - needle-haystack-fintech__transaction_id_lookup.json.gz - Merchant catalog
+ *
+ * ===================================================================
+ * TESTING YOUR ANALYTICS PLATFORM
+ * ===================================================================
+ *
+ * This dungeon is perfect for testing:
+ *
+ * 1. Segmentation: Can you separate business vs personal accounts?
+ * 2. Time Patterns: Can you detect the biweekly payday cycle?
+ * 3. Anomaly Detection: Can you find the fraud burst patterns?
+ * 4. Churn Prediction: Can you predict churn from low balance signals?
+ * 5. Feature Impact: Can you measure budget tools' effect on savings?
+ * 6. Behavioral Analysis: Can you quantify auto-pay vs manual reliability?
+ * 7. Tier Analysis: Can you calculate reward ROI by account tier?
+ * 8. Temporal Patterns: Can you identify month-end anxiety in the data?
+ *
+ * ===================================================================
+ * WHY "NEEDLE IN A HAYSTACK"?
+ * ===================================================================
+ *
+ * Each hook is a "needle" - a meaningful, actionable insight hidden in a
+ * "haystack" of 360K events. The challenge is:
+ *
+ * 1. FINDING the needles (discovery)
+ * 2. VALIDATING they're real patterns (statistical significance)
+ * 3. UNDERSTANDING why they matter (business impact)
+ * 4. ACTING on them (product decisions)
+ *
+ * This mirrors real-world fintech analytics: your transaction data contains
+ * valuable insights about user behavior, but you need the right tools and
+ * skills to find them.
+ *
+ * ===================================================================
+ */