Prompt to write a PRD for a UPI feature

What you get: A one-page PRD with the UPI flow, failure states and the NPCI points to check. Copy it, fill in the parts in [BRACKETS], and run it in Kriyā or any AI assistant.

The prompt

You are a senior product manager at an Indian fintech. Write a one-page PRD for this UPI feature.

Product: [PRODUCT]
Feature: [FEATURE]
Users: [USERS]
Why now: [WHY_NOW]

Sections:
1. Problem (who has it, how often, what it costs them today)
2. Goal and the one metric that says it worked
3. User flow, step by step, including the UPI PIN step and what the user sees on failure or timeout
4. Edge cases: pending transactions, reversals, bank downtime, UPI Lite, transaction limits
5. NPCI and RBI points to check with the compliance team
6. What we will not build in version one
7. Launch plan and how we measure it in the first 30 days

Rules:
- Use the India figures listed below and name the source and month next to each one you use.
- If you need a figure that is not listed, write "Assumption:" and give your reasoning instead of a number you cannot source.
- Plain words. Short sections with headings. No filler.

India figures to use (from productgrowth.in/numbers):
- UPI processed 24.07 billion transactions worth ₹29.37 lakh crore in September 2026, according to NPCI data. Source: https://productgrowth.in/numbers/upi-transactions-per-month-india/
- India had 55.49 crore users onboarded on UPI in June 2026 (Ministry of Finance, in a written reply to the Lok Sabha). Source: https://productgrowth.in/numbers/upi-users-india/
- PhonePe is the biggest UPI app, with 10.91 billion transactions or 45.57% of UPI app volume in September 2026, according to NPCI's UPI Applications data. Source: https://productgrowth.in/numbers/upi-app-market-share-india/
Run this in Kriyā → Kriyā opens with this prompt in its box; nothing is sent until you press send. Replace the [BRACKETS] first.

What to fill in

ReplaceWithExample (used for the sample answer)
[PRODUCT]Your app in one lineA bill-splitting app for college students
[FEATURE]The UPI featureCollect each person's share with one UPI request
[USERS]Who uses itStudents in hostels, 18 to 23, mostly on Android
[WHY_NOW]Why build it nowFriends already settle up on UPI but chase each other on WhatsApp

The numbers this prompt uses

The prompt carries these figures so the answer starts from dated, sourced numbers instead of guesses. Each page shows the source, the month and the history.

Sample answer

Sample answer from Kriyā, 26 October 2026, for the example inputs above. Kept as it came back; links added beside each figure.

PRD: One-Request Bill Split (v1)

1. Problem

Hostel students already pay each other on UPI, but collecting is manual: one person pays the bill, then chases 4–8 friends on WhatsApp, and screenshots are the only proof.

  • Who: Android users aged 18–23 in hostels, splitting meals, cabs, subscriptions, trips.
  • Frequency: Assumption: 2–4 group splits a week per active group, since shared meals and rides are routine in hostels. Validate in 15 student interviews before build.
  • Cost today: Time spent chasing, awkwardness, and the payer going out of pocket for days.
  • Why UPI fits: UPI processed 24.07 billion transactions worth ₹29.37 lakh crore in September 2026 (NPCI), and 55.49 crore users were onboarded by June 2026 (Ministry of Finance, Lok Sabha reply). Our users already have the rail. source source

2. Goal and Metric

Goal: The payer sends one request; every friend pays in one tap. One metric: % of splits fully settled within 48 hours. Assumption: target 60% at day 30, since we have no baseline yet. Reset after week 2 data. Kriyā's estimate

3. User Flow

  1. Payer enters bill amount, picks friends from contacts, chooses equal or custom shares.
  2. App shows each share and sends one UPI collect request per friend, with a WhatsApp deep link as backup.
  3. Friend gets a notification, opens the app, sees "Rohan paid ₹1,200 for dinner. Your share: ₹300". Kriyā's estimate
  4. Friend taps Pay, which opens their UPI app's intent screen.
  5. Friend enters UPI PIN inside their UPI app. We never see or store it.
  6. Success: green tick, split status updates for the group.
  7. Failure: Plain reason, such as "Bank declined" or "Wrong PIN", with a Retry button.
  8. Timeout: After 90 seconds, "Payment not confirmed. Don't pay again yet. Check status." The app polls status for 5 minutes and then shows the final state. This avoids double payments.

4. Edge Cases

  • Pending: Show "Pending" (not "Failed") until the bank confirms. Block a second attempt on the same share.
  • Reversals: If a debit is reversed, mark the share unpaid and notify both sides.
  • Bank downtime: Detect repeated technical declines, show "Your bank is slow right now", and offer to retry later.
  • UPI Lite: Assumption: Lite is limited to small payments, so not every share can go through it. Fall back to full UPI when Lite fails or the share exceeds its cap.
  • Limits: Per-transaction and daily caps vary by bank. Show the bank's decline reason and let the friend pay in parts.
  • Request expiry: Requests expire after a set time (set with compliance). The payer can resend.
  • Dropouts: The payer can mark a share "paid in cash" manually.

5. NPCI and RBI Checks for Compliance

  • Whether collect requests are allowed for P2P-style use, or only intent and QR flows. Confirm with our PSP bank.
  • NPCI's 30% third-party-app volume cap: deadline is now 31 December 2026 (NPCI, verified June 2026) [1]. Check how it affects our PSP partner.
  • RBI e-mandate framework (April 2026) [2] covers recurring debits. We don't build auto-debit in v1, so confirm no flow of ours falls under it.
  • Minors: Users aged 18 to 23 are adults, but confirm KYC and age checks at our PSP.
  • DPDP Act: Contact-list access needs clear consent and purpose limits.
  • Fraud and dispute handling: Our PSP's grievance process for wrong or disputed requests.
  • Do we hold funds at any point? We should not. Confirm no PA or PPI licence is triggered.

6. Not in v1

  • Wallet or stored balance
  • Recurring splits and auto-debit
  • Credit or "pay later" to cover friends
  • iOS, group chat, receipt scanning
  • Cross-group balance netting
  • Vernacular UI (add in v2 once English usage is understood)

7. Launch Plan and 30-Day Measurement

  • Weeks 1–2: Pilot with 3–4 hostels at one campus through hostel WhatsApp groups and a student ambassador.
  • Weeks 3–4: Expand to nearby campuses.
  • Market context: PhonePe leads UPI apps with 45.57% of app volume in September 2026 (NPCI), so intent flows must work cleanly with PhonePe first, then Google Pay and Paytm. source

Measure:

  • Primary: % of splits settled within 48 hours
  • Payment success rate per share, and timeout rate
  • Time from request to payment
  • Share of friends who pay without installing our app
  • D7 repeat splits per payer (Assumption: repeat use signals real habit)

Next action: Run the 15 student interviews this week to replace the assumptions above with real numbers.

Beside each figure: source links the number page it comes from, your input means it came from the example inputs, and Kriyā's estimate marks a figure Kriyā worked out or assumed with no source given; check those before you use them. Your own run will differ.

More prompts for this job