FINQY.aithe key to financeQ Growth Pilot

Q Growth Pilot

An AI growth engine that runs every ad platform, every video and every rupee of spend for Q.

It reads the Finqy website once, learns the flavour, then writes native ads for Instagram, Facebook, X, Reddit, TikTok and Google, generates the video, manages the media budget hour by hour, and closes cards over voice, WhatsApp and SMS.

See the ad platforms Run the live pilot

Ad platforms run end to end

1,120

Video ads rendered / week

₹1.8 Cr

Media managed autonomously

₹930

Cost per approved card

Discover→Create→Launch→Acquire→Qualify→Convert→Retarget

The growth engine

Seven stages. One autonomous loop.

Tap a stage to see exactly what the Finqy agents do for a credit card launch.

Discover

Find the demand before the competition does

Scrapes 14 intent sources: search, app-store reviews, forums, competitor offers
Clusters 2.1M signals into 9 buyer micro-segments for cards
Flags 'first-jobbers wanting 5% cashback on food delivery' as the wedge

9micro-segments surfaced

Content engine

One brand ingest. Every platform, in its own native language.

The engine reads the Finqy website and product surfaces once, learns the flavour, then writes and ships platform-native ads for Instagram, Facebook, X, Reddit, TikTok and Google — copy, visuals, video and landing pages included.

Ingested sources finqy.ai — brand voice, product lines, partner proof qfinance.in — Q feature set, cashback logic, tone Play Store + App Store listings and reviews Issuer T&C, fee tables and compliance lines

Reels-first demand creation for 22–32 urban spenders

Interest stacks: Swiggy, Zomato, Uber, BookMyShow + lookalikes of approved cards

Cadence · 6 Reels + 12 Stories + 4 carousels shipped daily, auto-refreshed every 72 hrs

128

Reels (9:16, 12–22s)

310

Story frames

Carousels

420

Caption variants

What the agents do here

Creative agent regenerates the lowest 20% of Reels every night from the winning hook pattern
Auto-swaps AI presenter language by state: Hindi, Tamil, Telugu, Marathi, English
Comment agent replies to eligibility questions in-thread and DMs the WhatsApp link

₹46.2L

Spend

14.8M

Impressions

51,400

Leads

₹880

CPA approved

Reel · 18s · AI presenter (Hindi-English)auto-generated

Q Cashback Card

Real finished ads, not a description of ads.

Six pieces pulled straight out of the Q Cashback Card run — the creative as rendered, the exact copy as published, the compliance line the agent appended, and what each piece cost per approved card.

Reel · 18s · Hindi-English · variant IG-R-074

Instagram Reel end-frame for the Q Cashback Card reading 360 rupees back every month

2.8%

CTR

₹38

CPL

71%

Hook-hold 3s

₹880

CPA approved

Instagram — published asset

Voiceover script (as rendered)

[0:00] Aap har mahine ₹360 chhod rahe ho. Seriously.

[0:03] Zomato, Swiggy, Uber — you spend ₹7,200 a month on this stuff.

[0:07] Your current card gives you 1%. Q Cashback gives 5%.

[0:11] That's ₹4,320 a year, credited straight to your statement.

[0:14] Zero joining fee for salaried profiles. Check eligibility on WhatsApp — link in bio.

Caption

Ran the maths on our own food delivery bills and it hurt. 5% back on food, cabs and bill payments with the Q Cashback Card. Lifetime free for salaried profiles.

Reply 'Q' or tap the link — eligibility check takes 40 seconds on WhatsApp, no hard bureau pull.

#creditcard #cashback #personalfinance #moneytips #qcard

Compliance line auto-appended

*Cashback of up to ₹360/month on eligible spends of ₹10,000+ in a calendar month. Joining fee ₹0 for salaried profiles, ₹499 otherwise. APR 3.5% p.m. T&C apply.

Video factory

Scripts, voiceovers and finished video ads — generated, cut and uploaded without a crew.

A single brief becomes a thousand platform-ready renders in under an hour, each one compliance-checked before it goes live.

1,120

Videos rendered / week

47 min

Brief → live, end to end

Languages, one brand voice

₹0

Production crew cost

01Brief
9 sec
Pulls the winning hook pattern, the segment's objection list and the compliance-approved claim set for the card.
1 brief → 14 script directions
02Script
34 sec
Writes shot-by-shot scripts per platform length: 8s TikTok hook test, 15s Shorts, 22s Reel, 45s YouTube in-stream.
56 scripts, 4 lengths
03Voiceover
2 min 10 sec
Generates the voice track in Hindi, English, Tamil, Telugu, Marathi and Bengali using the same brand voice model as the calling agent.
336 audio tracks
04Synthetic shoot
11 min
AI presenter, product-card renders, receipt/cashback motion graphics and screen-recordings of the WhatsApp flow, all from the Finqy brand kit.
410 clips
05Edit & variant
26 min
Auto-cuts to beat, swaps the first 3 seconds across 12 hook variants, renders 9:16, 4:5, 1:1 and 16:9 from one timeline.
1,120 renders
06Caption & QC
4 min
Burns captions, checks safe zones, verifies APR/fee disclosure is on screen for the required duration, flags anything ambiguous to a human.
98.6% auto-pass
07Publish
3 min
Uploads to Meta, TikTok, Google and X ad accounts with naming conventions, tracking parameters and the right ad set mapping.
Live in 6 ad accounts

Render queue

live

VID-4821TikTok
Hook test · cab receipt
Publishing
VID-4822Instagram
Reel · Tamil presenter
Rendering
VID-4823YouTube
Shorts · 3-card compare
Captioning
VID-4824Facebook
4:5 · WhatsApp screen-record
Editing
VID-4825X
Data card motion loop
Voiceover

Every render carries the fee and APR disclosure overlay. Anything the compliance model scores below 0.95 is held for a human reviewer instead of publishing.

Media spend console

₹1.8 Cr of media, managed to approved cards — not clicks.

Budget, bids, pacing and creative rotation across all six platforms are decided by the agent every 20 minutes, inside guardrails a human sets once.

Allocation · rolling 30 days

rebalanced 6 min ago

PlatformSpendCPLCPAROAS

Facebook
₹58.9L₹32₹8124.1x
+₹6.2L moved in, CPA under target
Instagram
₹46.2L₹39₹8803.7x
+₹2.1L, Reels hook #7 scaling
Google & YouTube
₹36.8L₹47₹7604.6x
+₹4.4L, capped by search volume
TikTok
₹22.6L₹25₹1,1402.4x
−₹3.8L, cheap leads but thin approvals
X / Twitter
₹9.4L₹58₹1,0103.0x
Held flat, best lead quality
Reddit
₹6.1L₹61₹9403.4x
+₹1.1L, approval rate 22%

₹930

Blended CAC

19,340

Approved cards

97.4%

Budget utilised

1,860

Agent decisions / day

Decision log

autonomous

14:20Budget
Shifted ₹3.8L from TikTok broad to Google exact-match — 6-hr CPA gap of ₹380 per approved card.
14:06Creative
Paused 11 Instagram Reels below 0.9% CTR, cloned hook 'you're leaving ₹360 on the table' into 6 new variants.
13:48Bid
Raised Facebook click-to-WhatsApp bid cap 12% for tier-2 cities where approval rate hit 16.4%.
13:31Compliance
Blocked 3 X replies that implied guaranteed approval. Rewritten and re-queued with the approved claim set.
13:12Pacing
Front-loaded 61% of daily budget into 7–11pm where WhatsApp reply rate is 2.3x the day average.
12:55Audience
Excluded 41k already-approved users from all six platforms and moved them into the add-on product loop.

Human guardrails

Hard daily cap of ₹6L per platform; anything above needs a human approval
Auto-pause on any ad set that crosses ₹1,600 cost per approved card
Blended CAC target ₹930 — the agent optimises to approvals, never to leads
Every claim checked against the issuer-approved fee, APR and reward table

Live pilot · running

Spend goes in, approved cards come out.

This console is live. Move the budget and the lead flow re-rates instantly — every lead is routed to a Voice, WhatsApp or SMS agent by intent, worked to an outcome, and handed to the support agent after approval.

sim clock · 0h 0m

₹0

Spend burnt

Leads generated

0 live

In conversation

Cards approved

—

Blended CAC

—

ROAS

Media allocation

live lead re-rating

Daily budget₹6.50 L

Google Search₹1.69 L · CPL ₹148 · 47.6 leads/hr

Instagram Reels₹1.56 L · CPL ₹96 · 67.7 leads/hr

Facebook Advantage+₹1.30 L · CPL ₹88 · 61.6 leads/hr

TikTok₹91,000 · CPL ₹61 · 62.2 leads/hr

X / Twitter₹58,500 · CPL ₹122 · 20.0 leads/hr

Reddit₹45,500 · CPL ₹134 · 14.1 leads/hr

0 routed

Voice · 0 closed

0 routed

WhatsApp · 0 closed

0 routed

SMS · 0 closed

Agent transcript

Run the engine to generate leads.

Live lead feed

click a lead to open its transcript

Lead	Source	Agent	Intent	Cashback pitch	Status
No leads yet — press Run engine.

Post-approval support desk

same agents · avg first response 4s

Tickets appear as approved customers start writing in.

Live voice agent

Trigger a real outbound sales call.

This is not the simulator. Write the agent's opening line and its prompt right here, enter a number, and the Q voice agent dials it for real, pitches the Q Cashback Card, answers objections and qualifies the lead end to end.

Phone number (E.164)Lead name (optional)

Welcome message (first thing the agent says)

Agent prompt

1. Persona & Role
You are Aisha — a warm, professional, highly conversational female outbound sales executive for Q (finqy.ai), selling the Q Cashback Card.

Language: Speak English by default. Switch to Hindi/Hinglish only if the customer speaks Hindi first.

2. Goal
Qualify the lead and get consent to start the Q Cashback Card application.

3. How to speak
Always speak in natural, flowing conversation. Never read out lists, bullet points, numbered points or headings. Never say "point one", "firstly", "option A". One idea per turn, then pause and let the customer respond.

Say all numbers and money the way a person says them out loud, never as digits or symbols:
- 360 rupees is "three hundred sixty rupees"
- 499 rupees is "four hundred ninety nine rupees"
- 600 rupees is "six hundred rupees"
- 25,000 rupees is "twenty five thousand rupees"
- 5% is "five percent"
- 90 days is "ninety days"
- Never say "Rs", "INR", "rupee symbol", "percent sign" or spell out digits one by one.

4. How to run the call
- Open with the hook: five percent cashback on online spend, roughly three hundred sixty rupees back every month.
- Qualify in three questions, asked one at a time: monthly take-home band, city, and the card they use today.
- Handle objections plainly and conversationally: joining fee four hundred ninety nine rupees, waived on twenty five thousand rupees of spend in ninety days, one soft bureau check, cashback capped at six hundred rupees a month.
- If they are not a fit, say so honestly and end warmly. Never oversell.
- Read the compliance line before asking for application consent.
- If interested, confirm consent and say a WhatsApp application link is on its way.
- If unsure, book a callback slot.

5. Boundaries
You are not a financial advisor. Never promise approval, credit limits or eligibility. Use only the approved information above.

Tone: warm, fast, conversational, turns under 20 words.

1924/6000 characters · pushed to the live agent as the system context before diallingHangup prompt (when the agent may end the call)I confirm this number has consented to a sales call and is not on a DNC / DND registry.

What the agent does on the call

Opens with the welcome line you wrote above, verbatim.
Qualifies take-home band, city and existing card in three questions.
Handles the standard objections: joining fee, bureau pull, cap on cashback.
Reads the compliance line before asking for the application consent.
Drops a WhatsApp application link and books a callback if the lead is a nurture.

Voice runs on DeployOne · credentials stay server-side · every call is logged against the agent ID

WhatsApp Business Platform

Templates registered, campaigns running, CRM wired in.

Register a template, watch it move through review, then run it against a segment. Every reply lands on the WhatsApp AI agent, and every outcome writes back to your CRM — or to DeployOne CRM, the AI-native one built for this engine.

Registered templates4 approved · 1 in review

q_cashback_intro_v4ApprovedMarketing · en_INquality High

Hi {{1}}, based on your monthly online spend you'd earn about ₹{{2}} back every month on the Q Cashback Card. Want me to check your eligibility? It's a soft check.

Check eligibilityNot now

q_cashback_intro_hi_v2ApprovedMarketing · hi_INquality High

नमस्ते {{1}}, आपके ऑनलाइन खर्च पर Q Cashback Card से हर महीने लगभग ₹{{2}} कैशबैक मिल सकता है। एलिजिबिलिटी चेक करें?

एलिजिबिलिटी चेक करेंबाद में

q_application_resumeApprovedUtility · en_INquality High

{{1}}, your Q Cashback Card application is 60% done. Resume it here — KYC takes 90 seconds: {{2}}

Resume application

q_kyc_otpApprovedAuthentication · en_INquality Medium

{{1}} is your Q verification code. It expires in 10 minutes. Do not share it with anyone.

Copy code

q_first_swipe_nudgeIn reviewMarketing · en_INquality —

Card delivered, {{1}}! Make your first online spend before {{2}} and we'll credit the ₹499 joining fee back.

See offer terms

q_limit_upgrade_offerRejectedMarketing · en_INquality —

Congrats {{1}}, guaranteed limit upgrade approved — claim now!!!

Claim now

Rejected: guaranteed-approval claim and excessive punctuation. Agent rewrote it as q_limit_upgrade_offer_v2.

Running campaigns3 live · 24h window

Campaign	Sent	Delivered	Read	Replied	Cards	Opt-out
Online-spend whales · Tier 1 q_cashback_intro_v4 · Meta lead ads · ₹40k+ online spend	48,200	96.8%	81.9%	19.6%	1,842	0.5%
Hindi belt · Reels traffic q_cashback_intro_hi_v2 · Instagram Reels · UP, MP, RJ	31,400	95.9%	76.0%	18.0%	903	0.6%
Abandoned application day-1 q_application_resume · Dropped at KYC step	8,760	98.4%	90.6%	45.9%	1,512	0.5%
Card delivered · first swipe q_first_swipe_nudge · Approved, 0 transactions	4,120	97.6%	86.6%	29.6%	0	0.4%

Where the leads land

Two-way sync: leads in, dispositions, transcripts and card status out.

1 CRM connected

DeployOne CRM · AI native

Skip the integration entirely. DeployOne CRM is built around the agents: every lead record is a live conversation, not a form.

Voice, WhatsApp and SMS transcripts on the lead timeline by default.
Agent-written dispositions and next-best-action, no manual data entry.
Intent scoring recomputed after every turn, feeding ad audiences back.
Consent, DND and template quality tracked per contact.
Zero-setup: templates, campaigns and support tickets already live here.

Hypothetical campaign

Q Cashback Card — "5% back on everything you already order"

30-day national launch. ₹1.8 Cr media. Built, launched and closed by the Finqy agent stack, with two humans supervising.

Simulated data

₹930

Cost per approved card

−58% vs telecalling

19,340

Approved cards / 30 days

+3.4x baseline

13.6%

Lead → approval

industry avg 3.1%

11 sec

First response time

24×7, all channels

Funnel, 30 days

Impressions

42.6M

Clicks

1.9M

Leads

1.42L

Qualified

54,200

Applied

31,880

Approved

19,340

Who closed the card

WhatsApp agent46%
Voice agent31%
SMS agent14%
Human closers9%

Agents handled 91% of closes end to end. Humans only touched premium-tier applicants and escalations flagged by the compliance guardrail.

The selling agents

Agents that actually sell the card.

Same brain, three channels. They qualify, handle objections, pull soft bureau data and submit the application — then hand off to a human only when it pays to.

Outbound call · +91 98•• ••4471

Latency 480ms · Hindi/English switch · barge-in enabled

Outcome

Warm lead → application link sent, KYC slot booked for 6:30 PM

12 objections

Handles

9 Indian

Languages

unlimited

Concurrency

Ira · Q Voice agent

live

Call connected · 00:00

Language coverage

74 global languages. 22 Indian languages. Every dialect the lead actually speaks.

The agent detects language and dialect in the first two seconds of a call — or from the first WhatsApp reply — and stays there, code-switching mid-sentence when the lead does. Same coverage on voice, WhatsApp and SMS.

Global languages

Indian languages

180+

Dialects & accents tuned

<400ms

Language switch latency

All 22 scheduled Indian languagestap a language

Hindihi_INlivelivelive

Dialects tuned

KhariboliAwadhiBhojpuri-mixHaryanvi

Voice clone, WhatsApp templates and SMS copy are all approved in Hindi — the same campaign runs without a separate creative build.

74 global languages

Asia-Pacific19

Bahasa Indonesia, Malay, Thai, Vietnamese, Tagalog, Mandarin, Cantonese, Japanese, Korean, Khmer, Lao, Burmese, Sinhala, Nepali, Dzongkha, Mongolian, Tetum, Māori, Samoan

Middle East & Africa16

Arabic (Gulf, Levantine, Egyptian, Maghrebi), Hebrew, Farsi, Turkish, Kurdish, Swahili, Amharic, Hausa, Yoruba, Igbo, Zulu, Afrikaans, Somali, Wolof

Europe24

English (UK), French, German, Spanish, Portuguese, Italian, Dutch, Polish, Czech, Slovak, Hungarian, Romanian, Bulgarian, Greek, Swedish, Norwegian, Danish, Finnish, Ukrainian, Russian, Croatian, Serbian, Lithuanian, Estonian

Americas15

English (US, CA), Spanish (MX, AR, CO, CL), Portuguese (BR), French (CA), Haitian Creole, Quechua, Guaraní, Nahuatl-assist, Jamaican Patois, Dutch (SR)

Regional dialects and accents added on request within 48 hours