Case studies

We keep score.
Here’s the scoreboard.

Every engagement starts with a baseline and ends with a before-and-after. Client names are anonymised where NDAs require; the numbers come straight from project reporting.

Healthcare · Kochi, Kerala · WhatsApp AI Agent

A 6-clinic dental group stopped losing after-hours patients

The situation

Six clinics, one overwhelmed front desk per location. 40% of WhatsApp enquiries arrived after 7pm and were answered the next morning — by which time a third had booked elsewhere. Reception spent two hours a day on the same twelve questions.

What we built

A Malayalam-and-English WhatsApp agent on the official Business API, trained on treatments, pricing and insurance policies. It books directly into each clinic’s calendar, sends prep instructions, and escalates emergencies to the on-call line immediately.

22s
median response time, from 4+ hours
+38%
appointments booked per month
11 hrs
front-desk time freed per clinic, weekly

“Patients now compliment our “night reception team”. There is no night reception team.”

— Group Operations Manager, Aster-affiliated dental group

Real Estate · Dubai, UAE · AI Sales Funnel

A Dubai brokerage cut cost per booked viewing by half

The situation

AED 90k/month on Meta and Google ads producing leads that sat in a shared inbox. Agents cherry-picked; average first contact took 9 hours; ad spend was judged on lead volume, not sales conversations.

What we built

Instant AI qualification in English and Arabic within 60 seconds of every ad click, budget and timeline scoring, automated WhatsApp nurture with project brochures, and direct booking into agents’ calendars — weighted to close rates.

−52%
cost per booked viewing
60s
first response to every lead
2.7×
viewings per agent per week

“The funnel qualifies harder than my best agent, and it never takes Friday off.”

— Managing Director, Off-plan property brokerage

Export / Trading · Kochi, Kerala · Workflow Automation

An exporter erased 30 hours a week of documentation work

The situation

Every shipment meant re-typing the same data into invoices, packing lists, certificates of origin and buyer emails — around 90 minutes per order, with costly errors reaching customs paperwork.

What we built

One order-entry form now drives everything: documents generated automatically, buyer notifications sent at each milestone, Tally entries posted, and a daily exceptions digest for anything needing human eyes.

30 hrs
documentation work removed weekly
−94%
document errors reaching customs
6 wks
to full payback on project fee

“We doubled shipment volume last season with the same back-office team.”

— Export Manager, Spice & food exporter

Construction · Riyadh, Saudi Arabia · Custom AI Automation

A Riyadh contractor turned 3-day tender reviews into 3 hours

The situation

Bidding meant senior estimators reading 200-page tender documents to extract scope, exclusions and deadlines — three days per bid, so the firm bid on a fraction of viable tenders.

What we built

A document-intelligence pipeline that ingests tender PDFs (Arabic and English), extracts scope items, flags risk clauses against the firm’s playbook, and produces a structured bid-review pack for estimators to verify.

3 hrs
tender review, down from 3 days
3.2×
more tenders bid per quarter
2
major projects won in first 90 days

“We now read every tender in the market. Before, we read the ones we had time for.”

— Head of Estimation, Interior fit-out contractor

Want a number like these on your board?

A 30-minute call is enough to tell you whether we can move your metric — and roughly what it would cost.

Book a free consultation