Repear

AI that calls repair shops for you to compare prices and availability.

Comment

Inspiration

Getting something repaired is still way more work than it should be.

You search for businesses, call them one by one, explain the same problem repeatedly, ask about prices and availability, and try to compare everything afterward. Most people aren't going to call 10 places just to find the best option. I know I wasn't.

So I started wondering why AI couldn't just make those calls for me.

That's what became repear.

What it does

repear gives you Peary, an AI agent that handles the calling for you.

Show Peary what broke, explain what you need, and repear intelligently finds businesses that match the job. Peary then makes real phone calls to real businesses, talks with them, asks the questions that matter, and brings the answers back to you.

The idea is simple.

One request. Peary makes the calls.

How we built it

I built repear primarily using Replit, connecting OpenAI, Vapi, Google Places, and other services into a React Native and Expo app.

The difficult part wasn't just making an AI talk on the phone. repear has to understand the request, figure out what type of business can actually help, find good matches nearby, make the calls, navigate real conversations, and turn those conversations into something useful inside the app.

Somewhere along the way, I also spent about $2,000 that could've gone into my daughter's college fund.

Thanks Aspen. It's for the best.

Challenges we ran into

Real phone calls are chaotic.

Businesses have IVRs. People put you on hold. Calls go to voicemail. People interrupt the agent. Some businesses give prices over the phone while others need to see the item first.

Getting Peary to handle those situations naturally without repeating himself, getting stuck in loops, or forgetting what the user actually asked for took a lot of iteration.

Then came everything involved with turning the prototype into a real iOS app. Authentication, production bugs, API costs, edge cases, App Store requirements, and several rounds of review.

Accomplishments that we're proud of

The biggest accomplishment is that repear actually shipped.

It's now live on the App Store and people other than my family can actually use something that started as a random idea.

The launch also gave me my first real signal that the idea resonated with people. My build-in-public launch post reached roughly 47,000 views, generated more than 100 upvotes and 130 comments, and helped put the project in front of thousands of people.

Since launching, repear has already reached hundreds of customers according to RevenueCat.

For a project I built myself, seeing Peary go from an idea to making real calls for real users has been pretty surreal.

What we learned

The biggest thing I learned is that there's a massive difference between building a chatbot and building an agent.

A chatbot can tell you what to do.

An agent has to actually go out into the real world and do it.

Once AI starts interacting with real people and businesses, everything becomes unpredictable. Making that experience reliable is much harder than making a good demo.

I also learned that shipping the app isn't the finish line. Turns out after you build the thing, you have to convince people to use it too.

Unfortunately, I have now discovered marketing.

What's next for Repear

Repair is the first problem I'm tackling, but the bigger vision for Peary goes much further.

There are still countless things that require making a phone call. Scheduling a dentist appointment, ordering a pizza, getting a quote, checking whether something is in stock, making a reservation, or dealing with a business that simply doesn't have a good online system.

Repair gives me one painful problem to focus on while building the technology behind something much broader.

Long term, I want Peary to become the person you hand those calls to.

If you don't want to make the call, ask Peary.

Built With

Share this project:

Updates

Submission history