Muhammad Sheraz

Senior Full-Stack Mobile Engineer

// one of Pakistan's early Flutter adopters · shipping Flutter since October 2019, < 1 yr after Flutter 1.0

I build complete products — a Flutter app for iOS and Android, a Next.js web surface, and the backend and CI/CD behind both — plus LLM features and AI agents.

See the work Read my resume

Islamabad, Pakistan · UTC+5

1class Sheraz extends Engineer {2  final role    = 'Senior Mobile App Engineer';3  final based   = 'Islamabad, Pakistan';4  final years   = 7;  // Oct 2019 → now5  final shipped = 150;  // apps built6  final ledUpTo = 20;  // developers78  @override9  Product build(Idea idea) => Platform(10        app: flutter(),11        web: nextJs(),12        api: backend(),13        ai: agents(),14      );15}

1# stack.yaml — what I reach for, layer by layer2name: sheraz3description: Mobile, web and backend as one product.45dependencies:6  mobile:   [flutter, dart, bloc, riverpod]7  web:      [next.js, typescript, react, tailwind]8  backend:  [node.js, firebase, supabase, postgresql]9  ai:       [openai, gemini, retrieval, tool-use]10  delivery: [github-actions, ci/cd, crashlytics]1112published:13  pub.dev: 3  # verified publisher sherazi.dev

1// agents.ts — automation that checks its own work2import { agent, loop } from "./runtime";34const research  = agent("research", [search]);5const plan      = agent("plan", [research]);6const implement = agent("implement", [editCode]);7const verify    = agent("verify", [runTests]);89export const run = loop(10  research, plan, implement, verify,11).until((r) => r.verified, { max: 3 })12 .then(analyzeResults); // what changed & why

// apps built: 150+
// years of Flutter · since 2019: 7
// largest team led: 20
// packages on pub.dev: 3

flutter
dart
firebase
next.js
typescript
node.js
supabase
postgresql
openai
gemini
riverpod
bloc
getx
tailwind css
mongodb
socket.io
stripe
google maps
github actions
vercel

import 'package:flutter/material.dart';
await supabase.from("orders").select();
final cart = context.watch<CartBloc>();
export default async function Page() {}
FirebaseFirestore.instance.collection('users');
const app = express().use(auth);
await agent("verify").run(task);
- run: flutter build appbundle --release
const reply = await llm.stream({ tools });
runApp(const ProviderScope(child: App()));

import 'package:flutter/material.dart';
await supabase.from("orders").select();
final cart = context.watch<CartBloc>();
export default async function Page() {}
FirebaseFirestore.instance.collection('users');
const app = express().use(auth);
await agent("verify").run(task);
- run: flutter build appbundle --release
const reply = await llm.stream({ tools });
runApp(const ProviderScope(child: App()));

// selected work

Whole products, not just screens

All case studies

E-Commerce & Retail

POS & Inventory Platform

Full-stack POS & inventory platform with offline-first sync

A POS and inventory management platform: a Node.js/TypeScript REST API on PostgreSQL paired with a Flutter app that keeps working offline and syncs when connectivity returns.

FlutterTypeScriptNode.jsPostgreSQL+3

Marketplace & Booking

Quick Rent

A rental-only listings platform rebuilt around a branch-based, role-scoped backend

A branch-managed rental marketplace connecting tenants with branch offices — a Next.js web platform plus a Flutter mobile app sharing one Supabase backend, with real-time messaging, commission tracking and full audit logging behind role-scoped admin and branch panels.

Next.jsTypeScriptTailwind CSSFlutter+4

Social & Media

Gage

Social sports platform — live NFL/NBA scores, friendly wagers & group chat

A feature-dense social sports app where friends track live NFL/NBA scores and news, challenge each other with 'gages', climb leaderboards and talk trash in group chat — with push notifications and deep links throughout.

FlutterFirebaseTypeScript Cloud FunctionsFirestore+2

// swipe →

// ai & automation

Agents that check their own work

Case studies on this site include LLM features — assistants grounded in their own content, retrieval over a library, streaming chat with live tool calls. The agents I build run as a loop: they research, plan, implement, then verify and analyze their own results before a person signs off.

OpenAI
Gemini
retrieval
tool-use agents
streaming UIs
agent loops

// see it in the case studies

.github/workflows/agents.ymlon: task

research-agent
Gathers context first — docs and the data a task touches.
planneeds: research-agent
Turns findings into steps, each with a check that proves it's done.
implementneeds: plan
Makes the change through typed tool calls — code, APIs, data.
verifyneeds: implement
Tests the result against the plan. A failed check loops back, not forward.
on failure → back to plan, retry
analyze-resultsneeds: verify
Reports what changed, why, and what a human should look at next.

$agents run --until verified

// career

git log --graph

Full resume

~/careeronmain$ git log --graph --decorate

3b05c43(HEAD → main)Feb 2026 – Present · 7 mos
Senior Mobile App Engineer @ Global Financial Media (GFM)
Cross-platform Flutter development.
+flutter+dart
1792931Feb 2024 – Feb 2026 · 2 yrs
Senior Flutter Developer @ Techanion
Led development of cross-platform Flutter apps for Android, iOS and web, from requirements through store release.
+flutter+dart+firebase+rest-apis+ci/cd
b923dceFeb 2021 – Feb 2024 · 3 yrs
Senior Flutter Developer @ Techorra
Led development and project management across multiple client apps, with code reviews and shared best practices.
+flutter+dart+firebase+cloud-functions+getx+bloc
1ebd95fOct 2019 – Feb 2021 · 1 yr 4 mos
Junior Flutter Developer @ GlowSIMS
Taught myself Flutter end to end and shipped production apps from the first year.
+flutter+dart+rest-apis