-
Cedric Roberge, a college student, built Pep AI — a peptide safety tracking app — and generated $50,000 in total revenue within 7 weeks of launch by identifying a trending niche on social media, building quickly with AI tools, and leveraging influencer marketing to drive organic growth.
-
Products and Offerings
- Pep AI is an iOS app that helps people track peptide usage safely, targeting users taking peptides for weight loss, skincare, tanning, and injury recovery
- Core features include a dosage calculator, peptide logging with injection site tracking, a research library sourced from PubMed, educational quizzes styled like Duolingo, and a lifestyle section with meal scanning, nutrition tracking, weight tracking, Apple Health integration, and progress photos
- Monetized through a $10/month subscription or $45/year subscription, both with a 3-day free trial, managed via RevenueCat
-
Metrics and Financials
- $51,000 total revenue in approximately 7 weeks
- $11,000 monthly recurring revenue (MRR) at the time of recording
- Nearly 2,000 active subscriptions in the trailing 28-day window
- Approximately 33,000 in trailing 28-day revenue
- One influencer Instagram post drove an estimated $10,000 in total revenue; a single story post generated $1,000; the first full post from that influencer received 50,000 views and produced roughly $4,000 in one day
- Cedric’s goal is to scale to $1 million per month
-
Strategy and Growth
- Primary growth engine is organic discovery driven by the trending topic of peptides — users are actively searching the App Store for peptide-related apps
- Built on a growing trend before it fully saturated, giving the app significant free visibility
- Pre-launch validation included Reddit posts and a waitlist of 300 people before building
- Influencer marketing was the most impactful channel: Cedric identified niche influencers early, had them post about the app, and attributed the majority of early revenue to this strategy
- Competitor apps have since launched with similar names (e.g., “Peptide AI”), but Pep AI ranks as the top search result, capturing spillover traffic from the trend
-
Tech Stack and Infrastructure
- Initial prototyping done on Replit using vibe coding — describing what he wanted and letting the tool build it
- Claude used as a technical guide for setup tasks like RevenueCat (in-app subscriptions), Resend (email), and other integrations
- Firebase serves as the backend for user data storage
- Apple App Store approval was the hardest part of the process, with multiple rejections due to concerns about medical advice; Cedric studied how similar apps positioned themselves to get approved
- Used Apple’s expedite review process to speed up approvals from days down to roughly 2 hours
- Total build time from idea to working app was approximately 2 weeks, plus about 1 week for App Store approval
-
Lessons and Advice
- Finding the idea: Scroll TikTok and Instagram daily looking for topics that multiple creators are discussing obsessively — if it’s new and unfamiliar, it may be an untapped opportunity
- Validating the idea: Check if any competing apps exist — having a few competitors is actually a positive signal that demand exists, even if none have been successfully marketed
- Improving on competitors: Read every review of existing apps to find what users like, dislike, and wish for, then feed that into Claude or ChatGPT to generate your app concept
- Pre-launch momentum: Recruit niche influencers before launch to validate interest and build buzz; a single post from the right influencer can drive thousands in revenue
- Distribution first: Marketing and distribution matter more than the product itself — founders should have a clear plan for how they will market an app before investing time building it
- Persistence: Cedric’s biggest personal advice is to never give up on yourself, even when others dismiss your ideas; he credits his eventual success to relentless self-belief and refusal to stop iterating
- Build fast with AI: Vibe coding tools like Replit and AI assistants like Claude make it possible for non-technical founders to build and launch an app in weeks rather than months