Sugam PokharelFull-stack product builder for founders shipping paid softwareNepal

What lookscompleteusually isn't.

I design and ship complete products for founders and small teams — from the interface people touch to the accounts, payments and infrastructure that make it real. Based in Nepal, working worldwide.

Continue to meet the maker

02About

I build the parts that make a product real.

I'm Sugam Pokharel, an 18-year-old developer working across the whole product — the interface people touch, the systems behind it, and the deployment that keeps it useful once it is live.

I care about practical, fast and dependable work. The stack changes; the standard does not: make it clear, make it resilient, and finish the last mile.

03 / Selected work

Four systemsdisagreed with themselves.

Each one started as a problem of measurement, not of building: something the system reported was not what was actually true. The interesting part of each project is that gap, so that is what is written up here.

Production record04shipped systems

Deep dives02featured builds

Live now01public product

live202601

Toolorah

Browser-side tool platform with its own billing layer

A rack of 53 small utilities that run entirely in the browser — text, PDF, conversion, design and developer tools — behind accounts, tiers and paid plans.

  • Astro 6
  • TypeScript
  • Supabase
  • Paddle
  • Vercel

The hard part

Stripe does not operate in Nepal, so the payment layer had to be rebuilt

The obvious billing integration was unavailable: Stripe has no support for Nepal-based businesses. Toolorah runs on Paddle instead, as merchant of record — Paddle is the legal seller, handles global sales tax, and pays out. That changes the integration in ways the documentation does not lead with. Paddle’s list endpoints are not read-your-writes consistent, so two concurrent checkouts both see an empty customer list and both try to create; the loser is rejected with customer_already_exists and the winning id has to be recovered from the error body rather than treated as a failure. Its webhook signature window is five seconds, not Stripe’s three hundred, so a slow cold start can fail verification on a perfectly valid delivery — nothing may run before the unmarshal call.

Tools
53
Categories
8
Rendering
Hybrid SSR
Billing
Paddle MoR
Visit Toolorah

active202602

View Counter

Reconciling two views of the same data

Reads public Instagram reel view counts locally, then reconciles the logged-out pass against an authenticated one to recover the reels anonymous visitors cannot see.

Pass 1 — anonymous42%

Pass 2 — reconciled100%

The second pass is authenticated and fills only the gaps — usually one request.

  • Node.js
  • Playwright
  • Electron
  • PDFKit

The hard part

The hidden reels were the ones that mattered

A logged-out scrape looks complete and is not. Instagram hides some reels from anonymous visitors, and disproportionately the popular ones — in one account a single hidden reel was 58% of its total views. So the scanner runs two passes and reconciles them: an anonymous pass that carries no ban risk does the bulk of the work, and an authenticated pass fills only the gaps, usually in one request. If you are paying people from these numbers, the difference is the whole point.

Passes
2
Recovered
up to 58%
Runtime
Local only
Output
PDF report

active202603

PixeletShop

Storefront with live order analytics

A commerce front end with catalogue, cart and an admin surface that charts order and revenue movement as it happens.

  • React
  • Firebase
  • Recharts
  • Vite

The hard part

Two audiences, one data model

The shopper and the operator want opposite things from the same records — one wants a product, the other wants a trend. The build keeps a single Firestore shape and derives both surfaces from it, rather than maintaining a reporting copy that drifts out of sync with what customers actually see.

Surfaces
2
Data
Firestore
Charts
Recharts

archived202604

RideBooking

Four surfaces, one dispatch core

A ride-hailing system built as four coordinated clients — customer app, driver app, admin dashboard and backend — around a shared dispatch model.

  • Node.js
  • React
  • Mobile

The hard part

The hard part was never the map

Every surface has a different idea of what a ride means: a request, an assignment, a job, a row. Getting four clients to agree on one state machine — and to stay agreed when a driver goes offline mid-trip — was the actual engineering, and the reason the project is worth listing.

Clients
4
Shared core
1

The pattern across the work

Measure the gap. Reconcile the system. Ship the truth.See how I build

04 / Capabilities

One builder.The whole product.

I move through the layers that usually create handoffs — from the first interaction to the system behind it and the release people can use.

A complete path from idea to live product.

Design

Product interfaces

Clear, responsive surfaces that make complicated tools feel immediate.

  • Astro 6
  • React
  • TypeScript
  • Vite
  • HTML
  • CSS
  • JavaScript
Shipped proof53

browser-side utilities shipped in one platform

Connect

Systems & data

APIs, accounts and data paths that stay dependable behind the interface.

  • Node.js
  • Supabase
  • Firebase
  • Python
  • MERN
  • Playwright
Shipped proof2 passes

reconciled into one accurate reporting flow

Ship

Launch & revenue

Deployment, billing and the last mile between a working build and a real product.

  • Paddle
  • Vercel
  • Electron
  • PDFKit
  • Recharts
Shipped proofPaddle MoR

global billing engineered from Nepal

Tools change. The ability to connect the layers is the capability.

05 / 2020 — 2026

Six years, four shipped.

The line does not stop here. Every period below is pinned to it, and it draws as you read them.

  1. 2020 — 2022

    The Beginning

    Started exploring web development through HTML, CSS, and JavaScript, learning how the web works from the ground up. This period was about experimentation, curiosity, and building the foundation that would shape everything that followed.

    • HTML
    • CSS
    • JavaScript

  2. 2022 — 2024

    Going Deeper

    Moved beyond the fundamentals and began exploring modern development more seriously. Expanded into React, Astro, TypeScript, Python, and the MERN stack, while building smaller projects to turn concepts into practical experience and better understand how complete applications come together.

    • React
    • Astro
    • TypeScript
    • Python
    • MERN

  3. 2024 — 2026

    Building for the Real World

    Took development further through an Advanced Web Development course while focusing increasingly on real-world execution. Designed, developed, and shipped four production projects, gaining hands-on experience across the full journey from an initial idea to a finished product used in the real world.

    • Advanced Web Development
    • 4 production projects
    • Idea to product

  4. Present

    Still Learning. Still Building.

    I continue to explore new technologies, refine my craft, and take on more ambitious ideas. My focus has shifted beyond simply learning frameworks or tools toward building useful, polished, performant, and production-ready digital experiences.

    Keep learning, keep shipping, and make every project better than the last.

    • Learning
    • Shipping
    • Improving

06 / Contact

Available for the right project

Have something useful in mind?

Let's makeit real.

Tell me what you are building, who it is for and where it is stuck. A short, honest email is the best place to begin.

Tell me about your projectsugampokharel28@gmail.com

Nepal / Working worldwide

© 2026 Sugam Pokharel