Today I launched the MVP of my first app: Search For Me.
It's a smart PDF reader that understands the content of your files. You can open up one or several PDFs, ask a question, and the AI will read through all the files to find the answer.
I'm by no means an expert entrepreneur, but maybe someone less experienced than me can benefit from what I've learned:
I purposely searched for a problem that I already knew how to solve for two reasons:
I already knew the AI theory behind the solution, I already coded many neural networks, and I already knew how to make the app (mostly, not the details) and it still took twice the time I thought It would take (a little over two months).
Work hard, but don't beat yourself If you don't satisfy your imaginary deadlines. Making progress every day is enough.
This idea is the "scratch your own itch" kind of thing. I read and I search on PDFs a lot (researching). This app will save me precious time and effort in searching for concepts and ideas on PDFs. Just to be sure, I asked several people with an "early adopter profile", and everybody said to me: "Wow, awesome!" or "It would be useful". So I took the chance to build the MVP.
I spent a week thinking about how I can validate this idea without having to code the AI and the app. I couldn't figure it out. The closest thing that I came out with is an assistant (an actual person) that looks through your files.
So I rolled up my sleeves and started coding. If you can, try to avoid programming on the validation phase.
The idea isn't validated yet, I just released the MVP today! But I won't work on new features until I get real market validation. That is until I get a few dozen happy beta testers, and several users pay me for the next version of the app.
I have to confess that I knew that this is an important step, and I still kind of skip it. Yes, I did the landing page and put something out there on Reddit and Twitter, but that is it. And for that, now I have less than 10 beta testers to give me feedback.
In my mind, I had the idea that I should promote it when the MVP was ready because I will have something to show. But now I have the MVP and no users, so I still don't know If this is a valid idea to work on.
I should've been promoting the app more. That is why now my main goal is to promote the app and find out if it is worth the time to keep working on it.
I have a little notebook right beside my monitor. Every day I add a few things to do on the day or the next day. It is important to use it as a diary but only for work. Like a mechanical engineer of a factory would have.
This habit helped me:
This is the one thing that I think I did right. Before I started the development of the app, I ran through a few mental simulations of what I would do. And for every crucial step, I made a stand-alone project where I only worked to accomplish that step (usually one or two days per step).
This served me as a validation of my capacity for accomplishing the final goal before investing a lot of dev time. And by the time I started developing the app, I already solved all the problems ahead, making my days an effortless developing walk.
If you have any questions or just want to talk, don't hesitate to reach through my twitter or IH profile!
Have a great day!!
What do you think about the post? Was it useful? Is my English weird or hard to read?
Robertino I downloaded and signed up for Search for me. Some observations
Sorry man, my experience was not great. I dont think I will download an app.
I think if you give me an option to upload a file and you analyze if on the cloud, it would be great. I still dont know what your app does BTW.
The certificate must have an error, hence the installation problems. I'll look into everything you said. Thank you so much for your time!
Edit:
Everything but the second point shouldn't happen. The main use-case of the app is to find a concept/answer when there is too much to look in to. Like searching for a concept where the user doesn't know which of the 20 PDF has the answer.
In that case, a web app where the user needs to upload 20 files is less convenient than an application. Nevertheless, I understand your point of view.
Thank you again for your time and the feedback!
Terrific learning Robertino. I would have done an MVP where I would tell a bunch of realtors or lawyer I would summarize the documents they get in PDF. The upside is I will know the type of documents.
Thank you @mukundmohan. I entertained a similar MVP, but researching the subject I encountered that lawyers, on top of the specific law jargon, use common words with different meaning. Facing the daunting task of recollecting a dataset specific for lawyers, I move away from that project. Although, with time and experts it can be done!