Every WhatsApp lead answered, qualified and warm

Around 100 ad leads reach you each day. By the time a rep replies on WhatsApp, many have forgotten they signed up.

Regional online marketplace · SG and HKBuilt in-house by our founder
5qualifying questions asked automatically
3AI agents working in sequence
0cold leads reaching account managers
In their shoes · Sales reps and account managers

Before and after.

A Singapore marketplace was getting around 100 Facebook Ads leads a day, and all follow-up happened by hand on WhatsApp. Leads forgot they had signed up, and real prospects went cold before an account manager could reply.

  • Reps answering cold replies by hand
  • Same qualifying questions typed every time
  • Good prospects go cold before contact
How it works

What we built, step by step.

Step 1

Sort every reply

The assistant finds each lead in the sign-up sheet and reads their reply as cold, warm or unclear, then routes it to the right agent.

Step 2

Ask the 5 questions

A friendly sales assistant asks 5 qualifying questions one at a time, reading the whole chat first so it never repeats itself.

Step 3

Keep them warm

A separate sequence sends a welcome video on day 0 and a success story from a similar business on day 2, without double sends.

Step 4

Hand off to the account manager

Once qualified, the account manager gets a Slack alert, and the full chat history is already logged in the sheet.

5qualifying questions asked automatically
3AI agents working in sequence
0cold leads reaching account managers
The full story

From busywork to running on its own.

The problem

A Singapore marketplace was getting around 100 Facebook Ads leads a day, and all follow-up happened by hand on WhatsApp. Leads forgot they had signed up, and real prospects went cold before an account manager could reply.

What we built

We built a WhatsApp assistant with three AI agents: one sorts each reply as cold, warm or unclear, one asks 5 qualifying questions one at a time, and one answers common questions until handoff. A separate sequence sends a welcome video on day 0 and success stories matched to the lead's service on day 2.

What changed

Every lead gets an instant reply, at 2 PM or 2 AM. Reps only talk to leads who have been through all 5 questions, and every conversation is logged so they have context before the first call.

Built withn8nWhatsApp BusinessOpenAIGoogle SheetsSlack

Built in-house by our founder, before or alongside Workchestra. Names and numbers in the screens are samples; the results are real.

Want this for your team?

A free 20-minute call with the founder. We will tell you honestly if it is worth automating.

Book a free call
Case studies

Keep exploring the work.

18 case studies from real teams. Each one started with a team doing the same task by hand, every day.

Case study 01Workchestra client

More than 1,200 candidates are routed and traceable from first contact to orientation, with nothing retyped by hand.

Philippine manpower agencyRecruiters and the agency's management1,200+candidates routed and tracked
Case study 02Workchestra client

Management can compare any period, refreshed every hour, while the operations team keeps the same workbook.

Philippine manpower agencyManagement at a manpower agencyHourlyrefresh, fully unattended
Case study 03

Response time dropped from hours to under 60 seconds, with no manual data entry.

Regional online marketplaceAccount managers and their sales lead< 60 sfrom ad form to CRM record
Case study 04

Every thread, meeting and client session now has one source of truth.

US executive coaching firmA CEO and his executive assistant3ranked time slots per request
Case study 05

802 at-risk accounts, 35% of the 2,267 in scope, got a timely personal message in a single run.

Regional online marketplaceAccount managers with hundreds of accounts802at-risk accounts messaged in one run
Case study 06

Research that took 20-30 minutes now finishes in under 2 minutes.

Regional online marketplaceAccount managers preparing for first calls< 2 minto research one company
Case study 07

Decisions take seconds because each card carries its own context, so anyone on the team can act.

US executive coaching firmThe CEO's team reviewing AI workHourlyAI runs reviewed in one place
Case study 08

3.04 TB and 36,118 files moved, preserving about seven years of client history in the firm's own storage.

US executive coaching firmA firm with years of session recordings3.04 TBmigrated end to end
Case study 10

Nobody pulls report data by hand anymore, and the team saves 8-10 hours a week.

Regional online marketplaceA marketing team running ads in two markets0manual data pulls
Case study 11

The mastersheet updates itself, so nobody runs VLOOKUPs.

Regional online marketplaceThe ops person who owns the mastersheet3data sources joined automatically
Case study 12

Coaches get overviews the same day and can turn one into a client email draft with a single button, now rolled out to every coach.

US executive coaching firmCoaches following up after client sessions8kinds of AI mistakes checked
Case study 13

A spoken note becomes a review-ready draft in under 30 seconds.

Internal teamAn executive who thinks faster than they type< 30 sfrom voice note to email draft
Case study 14

Both markets start the day with an enriched, cleaned lead list, and existing clients never reach sales.

Regional online marketplaceSalespeople who should be selling300leads discovered per run
Case study 15

Every submission is classified within seconds, before the sales team opens the file.

Regional online marketplaceSales reps calling new sign-ups< 2 sto verify each submission
Case study 16

A 2-day, 4-person process became one click.

Regional online marketplaceA sales manager approving monthly results1 clickto approve, down from 4 people and 2 days
Case study 17

Every request gets an answer within seconds and lands on the calendar, so nothing falls through the cracks.

Internal teamThe one builder everyone messages0requests left unacknowledged
Case study 18

The content team gets polished Google Docs drafts in both languages on schedule.

Regional online marketplaceA content team serving two markets2languages written at the same time