Why We Built Swop Labs: The Call That Went to Voicemail at 9pm

It Started with a Missed Call

The idea for Swop Labs did not come from a business plan competition or a venture capital pitch. It started with a phone call at 9pm that went to voicemail.

A friend who runs a specialist clinic in central Singapore told us about a pattern he kept noticing. Every morning, he would check his phone system and find five to eight missed calls from the previous evening. These were patients who wanted to book an appointment, ask about a procedure, or check insurance coverage. They called after their own workday ended, between 6pm and 9pm, because that was the first free moment they had.

By the time the clinic called back the next morning, most of those patients had already booked at another practice. They were not disloyal. They just needed an appointment, and the first clinic that answered got the booking.

The maths was brutal. Eight missed calls per evening, five evenings per week, at an average consultation value of SGD 200. Even at a 30 percent conversion rate, the clinic was losing over SGD 9,000 per month to a problem that had a clear solution: someone (or something) needed to answer the phone.

We Tried the Obvious Solutions First

The first attempt was a traditional answering service. It was better than voicemail, but the operators could not check availability or book appointments. They took messages, and by the time the clinic followed up, many callers had already booked elsewhere. The per-call pricing also made costs unpredictable.

The second attempt was an IVR menu. Patients hated it. Call abandonment rates went up, not down. "Press 1 for English, press 2 for appointments, press 3 for enquiries" is not the experience anyone wants when they are trying to see a doctor.

That is when we started building what became AI Front Desk. A voice AI that actually sounds like a competent receptionist, speaks English, Mandarin, and Bahasa, knows the clinic's services inside out, and books appointments in real time.

From One Clinic to a Product

The first version was messy. It was built for one specific clinic with hardcoded responses and limited language capability. But it worked. Within the first week, the clinic captured 23 after-hours bookings that would have been lost. The ROI was immediate and obvious.

The second clinic came through word of mouth. Then a restaurant owner who had similar problems during peak service hours. Then a home services company whose technicians could never answer the phone because they were always on-site.

Each new client taught us something. Restaurants need the AI to know the full menu and handle reservation modifications. Home services companies need emergency dispatch logic. Hotels need concierge-style responses. We built each capability and realised we were creating a platform, not a one-off tool.

Adding AI Search

The second product, AI Search, came from a different observation. While running a DTC ecommerce brand, we noticed that a growing number of customers were finding products through AI conversations rather than Google searches. "ChatGPT recommended your product" started appearing in post-purchase surveys.

We dug into why. The brand had inadvertently done several things right for AI visibility: comprehensive structured data, detailed product content, strong reviews, and third-party mentions on authoritative sites. These signals were enough for AI models to cite the brand in recommendations.

We formalised this into a five-dimension framework, tested it by deliberately optimising the brand's AI presence, and saw citations increase by 3.2 times in 90 days. The approach was repeatable, so we productised it as AI Search.

One Brand, Two Products, One Mission

Swop Labs is not trying to be an everything-AI company. We focus on two specific problems that we believe are the highest-ROI AI applications for Singapore SMBs.

AI Front Desk makes sure every call gets answered, in every language, at every hour. AI Search makes sure your business gets cited when potential customers ask AI models for recommendations.

Both products share the same underlying belief: your business is only as visible as the systems that represent it. If nobody answers the phone, you are invisible to callers. If AI models do not know about you, you are invisible to the growing number of people who discover businesses through AI.

We are built in Singapore for Singapore businesses. We price for SMBs. We support in English, Mandarin, and Bahasa. And we measure our success by our clients' revenue impact, not by feature count or funding rounds.

What Is Next

We are expanding AI Front Desk to cover more industries and more languages. We are building MCP servers that feed business data directly to AI models for more accurate citations. And we are working with the PSG grant framework to make our solutions as accessible as possible for Singapore SMBs.

If your business has a phone that rings and customers who search, we built Swop Labs for you.

Talk to Us

Chat with us on WhatsApp to start a conversation. We reply within one Singapore business day.