She Had a Baby, Used AI to Build the Nutrition App She Wanted

Lisa Lin had lengthy labored alongside coders. When she had her son, she tried it for herself.

The 33-year-old has labored in operations roles at a wide range of tech firms, together with Uber. However she was “non-technical,” as her business would name it; she could not code.

Then got here the AI revolution, and her child. Lin moved from a full-time operations position to a contractor whereas taking over the brand new job of stay-at-home mom.

She’d heard horror tales about these early months, however has been pleasantly shocked. “He is a contented child,” she mentioned.

When her child was about 5 months outdated, she started testing out dietary apps to introduce strong meals and establish allergens. She tried a well-liked diet app however discovered it overwhelming. “Everyday, it is a number of data to digest,” she mentioned.

All Lin needed was an app that might “inform me what I ought to do as we speak,” she mentioned.

“I am large on diet,” Lin mentioned. “I wish to ensure I am doing this proper as a result of, truthfully, my pediatrician did not inform me something. They only handed me a paper.”

So, she acquired out Lovable, an AI app builder, and began tinkering. She joined the rising class of “vibe coders” who assemble apps and websites with little-to-no engineering data. All they want is an AI software, some good prompts, a imaginative and prescient of what may assist make their lives higher, and a willingness to experiment.

Lin gathered knowledge on which 50-100 meals to first introduce to your child. Her data got here primarily from the Department of Agriculture’s database, with help from ChatGPT and a few on-line articles. She then constructed a every day guidelines with the objective of constructing positive her child was consuming the fitting meals on the proper time — and named it Nutribabe.

For a lot of vibe coders, the joys of prompting results in hours and hours of labor. Lin could not do this with a five-month-old; she hardly ever acquired an hour uninterrupted.

“I do not get to take a seat down on the pc for 3 hours and vibe code the whole lot,” Lin mentioned. “I haven’t got the posh to try this.”

Lin would usually spend thirty minutes with Lovable earlier than returning to her child when he cried or wanted feeding. Her peak hours had been within the morning, and it took about one month to construct the primary iteration.

Probably the most difficult process was constructing the app’s inner infrastructure. Nutribabe features a suggestion engine primarily based on the child’s age and preferences.

Having a techie husband helps. Lin met her accomplice in San Francisco, the place she moved after dwelling in China and Singapore. They matched on Coffee Meets Bagel and stayed in San Francisco earlier than transferring to Las Vegas.

Her husband is a machine studying engineer. “He is an professional on the whole lot AI,” Lin mentioned. “Typically, I get suggestions.”

That additionally helped inform her largest piece of recommendation.

“In the case of constructing as a non-technical particular person, a very powerful factor is getting suggestions,” Lin mentioned. She may simply vibe code a shoddy app, she mentioned, however would not know whether or not there have been bugs or safety considerations with out asking for assist from somebody within the know.

As soon as the app was constructed, Lin began utilizing it every single day.

“I positively felt extra organized,” she mentioned. “I felt extra assured and extra calm.”

Lin’s subsequent problem: changing the online app right into a cell platform. She needs to open it to the general public, however is planning to get in a bit extra person testing first.

“I nonetheless wish to validate some particular product options with my mother pal,” she mentioned.