IT · AI · Interior Design
Shah Interio — An AI front desk that never misses an enquiry
A design studio losing high-intent buyers inside a flood of inbound calls now qualifies 70% of enquiries before a human is involved.
Client overview
Shah Interio operates in interior design. The engagement ran across our IT · AI practices.
Problem statement
Shah Interio generated strong inbound demand and converted poorly. Sales consultants spent about 60% of their day repeating the same qualification questions, and roughly a third of calls arrived outside business hours and were never returned. There was no record of what had been discussed with whom.
Research
- Four hundred recorded calls and two months of WhatsApp threads were reviewed and categorised by intent, language and outcome.
- Only 18% of inbound conversations involved a buyer with a defined budget and timeline — the rest were price-checking or early browsing.
- Language analysis showed 54% of callers switched between Gujarati, Hindi and English within a single conversation, which ruled out single-language automation.
- Response-time analysis showed conversion collapsing after 20 minutes, and after-hours enquiries converting at near zero purely because nobody replied.
Planning and approach
- A voice agent handling inbound calls in all three languages with mid-sentence code switching, disclosed as AI at the start of every call.
- A parallel WhatsApp flow using the official Business API for enquiries arriving by message, sharing the same qualification logic.
- Scoring on budget band, timeline, property type and location, with high-intent conversations warm-transferred to a consultant immediately.
- Every conversation written to CRM with transcript, score and next action, so no context is lost at handoff.
Execution
Week 1 · Call design
Scripts, objection paths, escalation rules and disclosure language written with the sales head.
Weeks 2–3 · Build
Voice agent, WhatsApp flow, telephony routing and CRM write-back configured and tested internally.
Week 4 · Shadow pilot
20% of inbound traffic routed to the agent with every call human-reviewed and scored.
Weeks 5–6 · Scale and tune
Full traffic, weekly transcript review, refusal and escalation thresholds tightened.
Timeline
Week 1
Discovery and conversation design
Weeks 2–3
Development and internal testing
Week 4
Live pilot on partial traffic
Weeks 5–6
Full rollout and tuning
Month 8
Measured review against the pre-launch baseline
Results
Return on the engagement.
The agent costs less per month than a single junior consultant and handles a volume that would require four. The larger gain was conversion: qualified buyers now reach a human within minutes instead of the following morning.
3.6x
Return in eight months
70%
Enquiries handled autonomously
-50%
Conversations per closed deal
100%
After-hours enquiries answered
“I was sceptical about disclosing it as AI. It turned out customers preferred it — they ask blunter questions and waste less time. Our sales team only speaks to people who are ready to buy.”
Conclusion
What we would tell a similar business.
The gain here came from sequencing — fixing the measurable constraint first, then building around it. If your business shows the same symptoms Shah Interio did, the diagnosis usually takes one conversation and a look at ninety days of data.
More stories
Related case studies.
Next step
Tell us the problem. We will tell you what it takes.
A 30-minute consultation with the practice lead who would actually run your mandate — no sales layer in between.