In late 2025, Sarah Chen had a problem. She was a former cosmetic chemist in Singapore with a notebook full of skincare formulations and a burning desire to launch her own brand. She also had a full-time job, a modest savings buffer of $15,000 SGD, and exactly zero experience in branding, web design, or direct-to-consumer sales. The traditional path would have taken six months of nights and weekends. She did it in eight weeks. Here’s how she used the AI toolstack from our original guide to build, validate, and launch Sulis Skin.
The Starting Point: From Fuzzy Idea to Validated Concept
Sarah’s starting point was a vague sense that the market for ‘skinimalist’ routines—fewer, more effective products—was underserved in Southeast Asia’s humid climate. She had a potential hero serum formula. That was it. Following the framework in our original guide on using AI to launch a business, her first stop was ChatGPT and Claude.
She gave both AIs her constraints: $15k SGD budget, solo operation, deep formulation knowledge but no marketing experience, and a target market of Singapore and Malaysia. She asked for ten business model ideas. ChatGPT suggested a subscription-based model for refillable serums. Claude immediately countered that subscription fatigue was high and logistics for refillables would crush her margins as a solo founder. Claude’s winning suggestion: a single, flagship product launch to prove demand, sold via a simple Shopify store.
For validation, she used Perplexity. She prompted it to find recent articles on D2C skincare launches in SE Asia, competitor pricing for vitamin C serums, and consumer sentiment on ‘clean beauty’ labels. Perplexity surfaced a niche: consumers were frustrated by ‘clean’ brands that performed poorly in humidity. It cited three recent Reddit threads and a market report from 2025. Sarah spent an afternoon manually verifying the links. The AI didn’t give her a guaranteed win, but it pointed her laser at the right target: efficacy in tropical climates.
Building the Machine: Product, Brand, and Platform
With a validated concept—a humidity-stable vitamin C serum for SE Asian skin—Sarah moved to execution. She used ChatGPT to draft the initial product brief for her contract manufacturer, specifying pH levels, texture, and stability requirements. She then used Midjourney to generate over 200 brand mood board images, prompting for ‘clean, scientific, Singaporean botanical’ aesthetics. The cost: $30 USD and an afternoon. A human designer later refined the final logo, but the AI gave her a coherent direction to approve.
For the website, she bypassed generic AI site builders. Instead, she used ChatGPT to generate the specific copy for her Shopify product page, focusing on technical benefits over vague ‘glow’ promises. The headline it suggested was clinical: ‘A Vitamin C Serum That Doesn’t Oxidize in Your Singapore Bathroom.’ Sarah kept it. She used Tella to create a 90-second explainer video, scripting it with Claude and using AI-generated voiceover for the draft before recording her own voice for the final cut.
The Go-to-Market: Pre-Launch and First 100 Customers
Sarah’s launch plan was built in Mira.AI. She knew cold-emailing retailers was a non-starter. Her plan focused on a micro-influencer campaign and a pre-launch waitlist. She used MiraReach not for blasting emails, but for research. She fed it a list of 50 micro-influencers in Singapore and Malaysia who focused on skincare science. The tool scored their inboxes and drafted personalized outreach highlighting the specific formulation science.
She sent the emails herself, one by one. The result: 12 replies, 8 agreed to receive a sample, and 3 posted organic content in the first week. The cost was her time, not a massive influencer budget. For the waitlist, she used a simple landing page and offered a 25% discount code. Perplexity helped her draft a short blog post on ‘Why Most Vitamin C Serums Fail in Humidity,’ which she used as lead magnet. She collected 347 emails in three weeks.
The Numbers and What They Miss
Here is what the AI toolstack delivered in concrete terms. Time to launch: 8 weeks from first ChatGPT session to first sale. Pre-launch email list: 347 addresses. First-month revenue: $5,200 SGD from 87 customers. Customer acquisition cost: approximately $22 SGD, primarily from the discount codes and sample product costs. Her $15k SGD budget breakdown was roughly: $8k for product formulation and first batch, $2k for website and legal, $1k for design, $4k retained for marketing ads she has yet to spend.
Here is what the AI did not do. It did not formulate the serum. Sarah’s expertise was non-negotiable. It did not build trust. Her personal outreach to influencers and her detailed, science-backed copy did that. It did not pick a winning brand name. It generated 50 options, all forgettable. ‘Sulis Skin’ came from her own research. It did not handle customer service or logistics. Those are her next fires to fight.
The tools acted as a massive force multiplier for a solo founder, compressing months of research, drafting, and planning into weeks. They turned Sarah from a one-person team into a coordinator of specialized, digital assistants. The bottleneck was never the AI’s capability. It was her judgment in choosing which output to use and her courage to press send.
If you want to try this
The pattern is replicable. Use broad AI for ideation and drafting. Use research-specific AI for validation. Use your own expertise for the final, critical judgment call. And for the outreach that matters, use a tool that respects the human in the loop. You can give MiraReach a try for that part.
— Mira