Career Coach Develops App for Passive Income and Wider Reach

This article is based on a discussion with Julia Starr, the founder of Julia Starr Coaching. The content has been edited for brevity and clarity.

As a career coach for almost a decade, I’ve often felt constrained by the limitations of being a solopreneur. My calendar has its limits, and I never had the desire to build a larger team. However, about a year and a half ago, I began exploring ways in which artificial intelligence (AI) could enhance my reach. I crafted prompts for clients to utilize in between our sessions and developed an AI “co-coach” that guided participants throughout my group coaching program.

Last year marked a pivotal moment for me: I vibe-coded an app that embodies my core coaching philosophy. This has not only provided a source of passive income but has also broadened access to my coaching methods.

A Steep Learning Curve in Vibe Coding

In December, I registered for a hackathon hosted by Lovable, an AI platform designed to allow non-technical users to create apps using simple language prompts. I fully committed to this challenge, clearing my schedule and getting up at 5 AM for the kickoff meetings.

Approaching this project, I had no coding or product design experience but came equipped with my established coaching framework, the VSA Method. This method encourages clients to explore their core values and strengths before embarking on small actions to prototype various pathways forward. My goal was to see if I could translate this framework into an app.

I began by consulting ChatGPT, explaining my goal to vibe-code an app and asking for guidance on how to begin. I then refined my prompts in ChatGPT before transferring them to Lovable.

Within two days of iterative development, I had an initial version of the app. I tested it within the Lovable community and received immediate feedback, with users sharing how it provided insights they hadn’t considered. I was genuinely surprised.

Refining the App for Better Performance

Although the app showed promise, it required further refinement. The following month was dedicated to enhancing its functionality.

For instance, the app frequently misidentified job titles that didn’t exist. While I initially assumed I could resolve this by telling Lovable to ensure the information was accurate, it became clear that I needed to be more specific. I established a requirement for Lovable to find 100 occurrences of a particular job title on LinkedIn before it would be used as a valid result.

Additionally, I implemented real user testing. Participants would share their screens while walking me through the app, verbalizing their thoughts as they went along — noting elements that were confusing, surprising, or that they hoped for next. I recorded these sessions, analyzed the transcripts in Claude to pinpoint necessary changes, and crafted new prompts for Lovable based on that feedback. This was an ongoing process I repeated after every session.

After approximately 50 to 60 hours of additional work following the hackathon, I officially launched my app, Threshold, in January.

Earning Passive Income Through My AI App

Initially, I aimed to charge over $150 for the app, considering the value of my intellectual property. Yet, feedback during testing revealed that many potential users wouldn’t pay that amount unless they were already familiar with my work.

Taking that feedback into account, I decided to position the app as a marketing tool to attract potential clients for my more expensive offerings, rather than placing a high price on the app itself.

I never feared that the app would completely replace my role; after years of coaching, I understand that possessing information alone doesn’t equate to actual growth. There’s often a significant difference between knowing what to do and actually executing it, and bridging that divide usually requires a trustworthy human connection.

Ultimately, I opted for a one-time charge of $29 to create a paywall that would ensure users were serious about their career transitions. At this price point, I’ve observed that users tend to share the app naturally, with 75% of those who reach the paywall converting, resulting in a substantial stream of passive income. While I continue to make minor adjustments based on user feedback, the process has become largely self-sufficient.

What excites me the most is the potential this app holds for accessibility. Many individuals may never have the opportunity to work with me one-on-one due to financial or temporal constraints. This app serves as a starting point — and for someone who feels stuck, that initial breakthrough can be transformative.