Mayorkun
banner
samuelayan.bsky.social
Mayorkun
@samuelayan.bsky.social
63 followers 12 following 510 posts
Product Designer | Software Developer | Specializing in iOS and Web Applications | Shipped 10+ Products | AI-Expert | AI-Powered creator | Ad Strategist | Scale your brand into 7 figures Portfolio: https://mayorkun.framer.website/
Posts Media Videos Starter Packs
Pinned
I’m a product designer who codes—or a dev who designs. Either way, I love turning ideas into clean, intuitive iOS and web apps.

Over the last few years, I’ve shipped 10+ digital products from scratch using SwiftUI, React, and a whole lot of coffee.
Want to test the bot for your business?

👉🏽 Drop “PILOT” below or DM me.

Devs & founders I’ll post my GitHub + build breakdown soon.

#buildinpublic #AI #WhatsAppBot #SmallBusiness #Nigeria #NodeJS #Entrepreneurship
This bot won’t just respond.
It’ll convert.

It keeps customers engaged, collects orders fast, and frees owners to focus on what matters.

🚀 Next steps:

Finish order flow + DB

Launch 3 pilot tests (Lagos vendors first)

Public access later
I’m not building for tech bros.

I’m building for:
👗 Fashion vendors
🍰 Bakers
💇🏽‍♀️ Hairstylists
🍛 Restaurants
💼 Freelancers

The people already running their entire business on WhatsApp and just need a little automation help.
Next up:

- Order capture (collect name, item, quantity, address)

- Auto-save to database (MongoDB)

- Vendor dashboard

- AI that learns your brand’s tone + pricing style
My favorite moment last week:

“✅ WhatsApp client ready — session saved
📩 Incoming: “Hey bot, how are you?”
🤖 Replied: “I’m doing great! How can I help you today?”
What I’ve built so far 👇🏽

✅ WhatsApp client using Node.js

- AI replies (first Hugging Face, now Ollama, fully offline)

- Real-time message parsing & session saving

- Error logs + self-recovery when it breaks
So I decided to build something simple:

A bot that lives inside WhatsApp, answers questions, takes orders, and passes control back to the vendor when needed.

No app. No dashboard login. Just WhatsApp.
Most small business owners (especially here in 🇳🇬) run their entire business on WhatsApp:
messages, orders, deliveries, customer service…

But one problem keeps coming up 👇🏽

They can’t reply to everyone fast enough. Every “I’ll respond later” = a lost customer.
I’m building a WhatsApp AI Assistant from scratch for real vendors and small businesses, not tech demos.

It replies, takes orders, and hands off to the owner all inside WhatsApp.

Here’s how it started 👇🏽
When speed meets empathy, that’s when products feel alive.

What’s the one “works perfectly but feels wrong” moment you’ve had while building?

#ProductDesign #WebDev #iOS #AI #DesignThinking #DeveloperLife #BuildInPublic
That’s when it hit me that sometimes the bug isn’t in the code, it’s in the design decisions.

As devs, we debug code.
As designers, we debug feelings.

I’ve shipped 10+ products, but this balance still humbles me every time.
Have you ever built a built a feature that looked great in code, but felt off in use.
It passed every test, ran fast, and even had perfect CRUD logic…
but when you tried it as a user, it felt confusing.

Code: works perfectly Design: feels wrong
And if you want, I can share a small checklist I now run through before every release DM me or drop a comment.

#ProductDesign #WebDev #iOS #DeveloperLife #LessonsLearned
We all get burned by these “tiny mismatches”. If you’ve ever lost hours (or days) due to schema mismatch, contract mismatch, or dev/staging drift, would love to hear your worst story. Let’s laugh and learn together.
After fixing, the feature worked in staging, then production. Users got the updated CRUD flow without the nasty UI weirdness or missing data. And I got 2 new messages that week from people wanting help with exactly that kind of bug (developer to product designer overlap).
• Always sync schema changes & client contracts before deployment.
• Automate migrations / versioning.
• Add logs and tests specifically for field mismatch / contract changes.
• Do small, incremental changes instead of “big sweeps” in one commit.
“user_email” to “email” but forgot to update migrations. Also, the client-side cache was holding the old field name, so requests were silently failing.

The result? What should’ve taken 30 minutes stretched into 3 days. I lost sleep, annoyed teammates, and re-scoped meetings.
Everything looked good locally. The UI flows, the API endpoints, even the validation. But when I pushed to staging, the update operation broke. No errors in logs, no clear culprit.

After 3 hours of digging: the database schema in staging didn’t match dev. I had changed a field name locally
This is how my simple CRUD feature cost me 3 days and what it taught me about design, deployment & expectations

I shipped over 10 products, but sometimes the smallest things sting the hardest. This past week I built a CRUD interface for a side project (web + iOS) that seemed straightforward.
For real if I am not the one designing your product then you are wrong cos I just got another recommendation this morning.

I am super talented
Patiently waiting on when we can host spaces on this platform.

Who else is with me?
Ever had a small but fatal mismatch cost you hours? Drop your story below let’s laugh (or cry) together.

Or DM me if you want help adding timezone-safe schedulers to your app.
#WebDev #Automation #ProductDesign
That small mismatch meant 0 jobs executed.
Lesson learned: Always simulate your CRON in the target environment first.

The snippet I used to test the image in the this post.
Built an automation that emails users a daily digest. It worked locally. But when deployed, nothing ran.

Yesterday, I spent 4 hours debugging a CRON job that ran every 5 minutes only to realise the timezone was wrong.

Turns out: My CRON expression was set to UTC, but my server was using +1 (WAT).
Glads to be on this project. Happy Sunday people