Nylo: Your Second Memory

One home for everything you save.

Comment

Nylo: Your Second Memory

Nylo is a personal AI penguin manager that helps you save, find, plan, and actually finish the things you save online. Send Nylo reels, links, notes or messages on Instagram or WhatsApp. He remembers them, finds them when you ask, turns them into plans, and keeps you accountable until they're done.

Inspiration

It started with something I noticed on Instagram.

My girlfriend saves a lot of reels: recipes she wants to try, workouts she wants to start, places she wants to visit, side-hustle ideas. Then I realised I do the same thing. I save 10+ reels almost every day, because I'm sure I'll need them later. But I almost never find them again.

And it isn't just Instagram. Screenshots, WhatsApp messages, links, notes and random ideas are scattered across my phone. I usually remember what I saved, but not where I saved it.

So I started with building a screenshot organiser. Its reel got 19K views, then viewers fell off. People didn't see the vision.

I tried multiple variations of organisers, in different reels.

Then, on 8 September, I posted a simple question on Instagram: "Am I the only one who saves 10+ reels every day?" The reel and its trial versions reached around 200K views, and 383 people signed up to test within two days.

But the most important part wasn't the views or the sign-ups. It was the comments. People kept asking, "Can I save directly from Instagram?" Others wanted the same for WhatsApp.

The problem was "I save things everywhere, and I can never find them when I actually need them." So instead of another app where you organise everything yourself, I put Nylo where people already save things. That is how the screenshot organiser became Nylo.

The traction so far

  • 1,225 people signed up to test Nylo before it was even built
  • 842 people used Nylo in the first 7 days after the chats opened
  • 2,078 things saved, by testers in 38 countries
  • 323 people joined the WhatsApp community
  • @shreyanbuilds grew from 0 to 1,045 followers and @heynylo from 0 to 1,126, in 50 days
  • The main Nylo launch reel reached ~100K views, with 2,110 likes and 391 comments

  • Nylo is live on the App Store

  • Get the app: https://apps.apple.com/app/id6805359169

  • Try the chat: send anything to @heynylo on Instagram, or message Nylo on WhatsApp at https://wa.me/919286650640

  • Built in public: https://www.instagram.com/shreyanbuilds/

  • Bundle ID: com.shreyannarula.shots

The four categories I am entering

BuildInPublic. I built Nylo in public on Instagram, starting from zero followers. I used reels to test ideas and see what people actually wanted, and one experiment changed the whole product: 1,225 people signed up to test Nylo before it was even built, and 323 joined its WhatsApp community.

Productivity. Nylo is a personal AI manager for everything you save. Send it from where you already are, find it again by asking, and turn it into a plan you actually finish. Save it. Find it. Plan it. Finish it.

Grand Prize. 842 people used Nylo in the 7 days after the chats opened, the tester list grew to 1,225 people across 38 countries, 2,078 things have been saved, and Nylo is live on the App Store.

Next Gen. A real, live AI product built by me, and I'm a student at IIT Roorkee. It brings AI into the apps people already use every day, Instagram and WhatsApp, instead of asking them to change their behaviour.

Pricing

The free version keeps 100 saves a month, with manual buckets. Nylo Pro adds asking Nylo questions, automatic planning and unlimited saves, for $9.99/month or $39.99/year, with a 3-day free trial. Subscriptions are handled through RevenueCat.

Why Nylo is different

  • It lives where people already save things.
  • It finds answers from your own saves.
  • It turns saving into doing and implementing.

How I built it

  • iPhone app in SwiftUI with three tabs (Plan, Chat, Saves), sign in with Apple, Google or email, and an onboarding chat that shows you exactly what to send Nylo.
  • RevenueCat handles pricing, offerings, trials and restoring purchases. The server checks the same Pro entitlement, so buying Pro in the app also unlocks Pro when you use Nylo on WhatsApp.
  • The chats run on a Cloudflare Worker using Meta's official WhatsApp Business Cloud API and Instagram Messaging API. Nylo only receives what people actively send to its accounts.
  • Saves live in Cloudflare D1 and R2. WhatsApp saves are encrypted at rest, with the key kept separately from the database.
  • Search without a vector database. Nylo embeds your saves, finds the closest matches by meaning in plain JavaScript inside the Worker, re-ranks them, and has a small model answer only from those sources, telling you which saves it used. 189 questions have cost only a few cents so far.
  • Reminders use Apple's push notification service, plus local notifications for the morning plan and evening check-in.
  • A web account at app.heynylo.com, and an admin portal where I see testers, usage and every piece of feedback in one place.

What testers changed

I keep all tester feedback on one board, with who said it and where: the WhatsApp community, TestFlight, the in-app board and App Review. It has 30 cards so far. The product itself changed because of it: people asked to save directly from Instagram and WhatsApp, and asked for a WhatsApp community. Both are now part of Nylo.

"I would like the add note feature on every saved item." Tester on TestFlight. Notes on saves are now in build 17.

"I think there should be an option to delete selected items." Tester in the WhatsApp community. Selected-item deletion shipped in build 17.

"The AI is not able to see what is in the reel." Tester in the WhatsApp community. Reading what is said inside a reel is next.

I wanted early testers to feel they were helping build something they'd actually use: the first 20 testers got 6 months of Pro free, the next 40 got 4 months, and anyone who joins now gets 1 month.

Who wants it?

My tester list mirrors where my Instagram audience comes from: 80.9% India (vs 82.9% of my audience) and 6.0% Pakistan (vs 6.3%). Outside India, 54% of testers use iPhone, against 32% in India, including 63% in the US, 86% in Canada and 93% in the UAE. For a subscription iPhone app, that shows where paying users are likely to be.

Challenges I ran into

  • App Store rejection. Apple rejected v1.0 over the wording on a permission screen. I fixed it and resubmitted in about a day and a half.
  • Generative AI rules. Apple rejected v1.1 for the China mainland storefront (As I used OpenAI for the AI features). I removed that storefront rather than hide which AI Nylo uses.
  • WhatsApp ban. WhatsApp banned my number for most of a day after I added testers to the community by hand. I ran the community from a friend's phone until my account was back.
  • Building during exams. My university exams were 10 to 12 September, right after the reel that changed Nylo's direction.

What I learned

  • People will tell you what to build, if you actually listen. The comments on one reel turned a screenshot organiser into something people use straight from Instagram and WhatsApp.
  • Make the action obvious. My first public vote got zero responses because I never clearly told people what to do. Now every call to action is specific.
  • Content needs one clear idea. A weekly recap covering several topics got ~300 views, against my usual ~2,000.
  • Build, test and change quickly. The original plan was not the final product, and that became the most important part of building Nylo.

What's next for Nylo

The next updates come straight from what testers asked for:

  • Understanding what's said inside reels, not just their captions
  • New chats with conversation history, so you can continue where you left off
  • Full-size iPad support

Beyond the product, I'll make more content for the US, Canada and the UAE, where a bigger share of my testers already use iPhone. The goal is simple: next up is the first paying user.

Built by Shreyan Narula, student at IIT Roorkee.

Built With

Share this project:

Updates

Submission history