By playing, you agree to uListen's Terms and Privacy Policy and YouTube's Terms of Service.

The Calum Johnson Show

The Vibe Coding Expert: The Fastest Way To Build & Launch $1M Al App SOLO (Claude + ChatGPT)

Timestamps About the Guest / About the Episode Chris Raroque is a self-made entrepreneur behind Ellie (daily planning), Luna (budgeting), and Mogul (personal CRM)—apps used by thousands every day, with 50,000+ signups and 1,000+ paying…

Chris RaroqueguestCalum Johnsonhost

Aug 25, 20251h 52mWatch on YouTube

Summaries, chapters, transcripts, and FAQs on this page are made by uListen, not YouTube.

Great conversations run long. Their best ideas don't have to.

uListen is free on iPhone, iPad, Android, Chrome, and Claude.

CHAPTERS

Jump to publisher timestamps

Chapter summaries

Why AI coding finally unlocks building for non-coders
Chris frames this as a unique moment where people without a traditional coding background can still build real products. He explains how AI tooling changes what’s feasible and why he wants to speak directly to beginners who feel “non-technical.”
- •Chris says building apps is more possible than ever, especially now with AI tooling
- •He describes entrepreneurship as repeated problem-solving and seeing everyday “pain points” as build opportunities
- •Calum sets the theme: a full blueprint from idea to brand to building with Claude Code
First dollar online: personal CRM, imposter syndrome, and learning to charge
Chris recounts building his first serious app—a personal CRM—and leaving it free for years despite thousands of users. He describes the mindset shift that happened when he finally charged and how that changed his confidence and approach to building.
- •Chris explains what a personal CRM is and why he built it for his own networking needs
- •He describes feeling embarrassed to charge and assuming “anyone could build this,” then realizing that wasn’t true
- •He emphasizes valuing your work and how charging reinforced that he was creating real value
Showing the receipts: Ellie, Luna, Mogul—and what “serious” AI-built apps look like
Chris shares his most notable apps and what drove each one, including usage and paying-user traction. He also pushes back on the idea that AI-built products must be “AI slop,” arguing you can build polished, impactful software with AI in the loop.
- •Ellie: built from a personal time-boxing need; thousands use it daily and it has paying users
- •Luna: manual budgeting by design to change spending habits; built to push design polish and interactions
- •Mogul: his earliest serious app and the first to make money
- •Chris argues AI coding can produce serious products, not just quick throwaway demos
What Claude Code is—and why Chris switched from Cursor
Chris explains Claude Code as a terminal-based AI coding tool that behaves more like an “agent.” He argues the key shift is that you can often operate without living in a code editor, which changes who can realistically build software.
- •Claude Code is terminal-based and can feel unfamiliar because it isn’t centered on a traditional code editor workflow
- •Chris calls it one of the closest “agentic” coding experiences he’s used
- •He says it’s the first time he’s seen a credible path where some users might not need to review code constantly
- •Calum presses for why “not seeing the code” is significant for beginners
Why Subscription Monster: picking a problem, defining constraints, and validating demand
Chris breaks down how Subscription Monster started from a concrete business pain point: not being able to quickly answer what software subscriptions cost across accounts. He emphasizes pairing the problem with constraints (time, goals) so projects actually ship.
- •Chris describes the trigger: needing an accurate subscription spend number and resorting to manual bank-statement review
- •He does quick searching to confirm the gap and treats constraints as a critical design input
- •Calum reinforces how lack of deadlines kills momentum and shipping
Idea triage: why products don’t exist, MVP scope, and when AI is the advantage
Chris explains how he filters ideas by asking why they don’t already exist and whether AI meaningfully changes what’s possible. He highlights that the best AI opportunities often live in tedious classification/cleanup problems that were previously expensive to solve.
- •He asks: why doesn’t this exist, and is the reason truly compelling or just a market blind spot?
- •He evaluates whether he can get to a meaningful MVP in weeks rather than years
- •He adds a newer filter: can AI itself solve the underlying hard part (like messy transaction labeling)?
- •He points out bank transaction naming inconsistencies as a key challenge that now maps well to AI
Using Claude Research as a market sanity check (and what it revealed)
Chris shows how he used Claude’s research mode to investigate why a “Rocket Money for business subscriptions” isn’t common. The outcome reshapes the opportunity: possibly too small for venture-scale, but attractive for a solo developer.
- •Chris uses Claude Research to gather and synthesize reasons the product category is rare
- •He highlights complexity in payment methods and transactions, plus practical constraints and regulations
- •He focuses on business reality: acquisition cost and market size may not justify a big company’s focus
- •He reframes: what’s small for a company can still be excellent for a solo builder’s constraints
The rough business plan: first dollar, revenue ceiling, and acquisition channels
Chris outlines the limited set of business-plan questions he actually cares about early: first-dollar path, likely revenue ceiling, marketing channels, and expansion options. He shares a channel mix and why direct outreach might work unusually well here.
- •Chris prioritizes: how to get the first dollar, revenue ceiling, marketing approach, and feature expansion paths
- •He estimates a solo ceiling and notes scaling beyond that likely requires more features
- •He names likely channels: build-in-public, SEO, Reddit, word of mouth, and cold email
- •He describes strong reactions from business owners as validation that outreach could convert
Naming and brand strategy: making “Subscription Monster” stick
Chris walks through how he uses AI to generate options, reject weak ones, and iterate toward a name that’s both memorable and searchable. He explains why including “subscription” matters for clarity and distribution, and how the name shaped the mascot direction.
- •Chris uses AI outputs as creative sparks, then relies on his own rejection filter to narrow down
- •He explores a meerkat concept, rejects it partly due to spelling/clarity issues, and pivots
- •He prioritizes having “subscription” in the name for clarity and SEO
- •He connects the name to an emotional story: subscriptions as a hidden “monster” businesses avoid confronting
Why mascots matter: trust, memorability, and product personality
Chris argues mascots create a distinctive identity people remember and emotionally connect with. He explains how mascots affect tone—especially for business software—and how they can improve UX through empty states and in-app visuals.
- •Chris describes how users became attached to the Luna mascot and how it personalized the product
- •He says mascots are memorable but risky if they’re too playful for the category
- •He explains mascots can elevate UX through empty states and repeated visual moments
- •Calum connects mascots to the broader question of branding and marketing beyond “just building”
Live mascot build workflow: hand-drawn base → AI variations → final logo asset
Chris demonstrates his iterative process for generating mascots: start from hand-drawn art, feed it into ChatGPT for stylistic continuity, then refine through many small changes. He shares practical tactics like restarting chats and avoiding overly large instruction bundles.
- •Chris shows examples of mascots and explains using hand-drawn art to achieve a more original, consistent style
- •He iterates from a ghost-style reference into a “Subscription Monster” direction, then refines tone for B2B trust
- •He emphasizes small, controlled edits over big multi-change prompts to avoid generic-looking outputs
- •He explains why he starts new chats to avoid the model overfitting to earlier generations
- •He notes the benefit of generating many mascot variations for UI empty states and product polish
Working prototype proof: Plaid-connected data, real transaction volume, and finding duplicates
Chris shows the Subscription Monster prototype running with a real connected business bank account and a large transaction history. He explains how quickly he reached a useful prototype and how the system helped surface redundant spending.
- •Chris demonstrates the app pulling and filtering transactions and subscription-related views
- •He mentions syncing and processing a large set of transactions and using the prototype to audit spend
- •He shares discovering duplicate spend (two accounts for the same tool) as a concrete “aha” result
- •He reveals he tracked the build via a structured prompt log and categorizes prompts by type
Claude Code 101: terminals, project setup, auto-accepting changes, and interface-first building
Chris explains how Claude Code is operated through a terminal, how he navigates into a project folder, and how permissions/auto-accept flow works. He then builds a tiny app from scratch and introduces his core build rule: perfect the interface first, then wire up functionality.
- •Chris defines terminal basics (e.g., changing directories) and how Claude Code runs inside a project folder
- •He shows Claude Code analyzing an existing codebase and summarizing the architecture
- •He spins up a new mini project, demonstrating beginner friction and how Claude can coach through it
- •He highlights a common beginner pitfall: doing too much in one prompt; recommends smaller steps
- •He explains the “interface-first” workflow: prototype UI with hardcoded data before adding real backend logic
Common beginner mistakes, step-zero marketing with waitlists, and compounding confidence
Chris closes with practical lessons: avoid oversized prompts, ask specific questions, and treat AI like a human collaborator. He explains why he builds waitlists early to gauge interest and recruit beta users, then reflects on how skills and confidence compound over time.
- •Chris says the biggest mistake is trying to do too much at once; he recommends iterative, specific instructions
- •He explains waitlists as early validation, a beta-user pool, and the on-ramp to first paying users
- •Calum frames waitlists as “step-zero marketing” that creates momentum before launch
- •Chris tells his past self it’s possible and that skills compound rapidly with repetition

Publisher timestamps

Get more out of YouTube videos.

Read along on your phone, search it beside the video, or ask your favorite AI app what was said.