Open to remote roles

AI products,Interactive systems,Full stack engineering

I turn complex systems into intuitive experiences.

Senior Full Stack Engineer building AI products, high-performance interactive interfaces, and scalable systems from concept to production.

Explore selected workLet’s build something
outputtrace(hasib)

Muttakin Islam Hasib

Based in Dhaka, Bangladesh, GMT+6, and open to remote roles. Every capability and project on this map traces back here.

Feeds

7+ years
shipping production software
4
countries of remote teams
Read about me

Select a node to trace what it’s built from and what it feeds. Drag nodes to rearrange; arrow keys move along the wires.Tap a node to trace what it’s built from and what it feeds.

Selected work

Five systems where the hard part sits below the interface, and the interface still has to make it feel simple.

ImagineArt Workflows: App Builder

ImagineArtSenior Frontend EngineerFeb 2026 – Sep 2026

Turning node-based AI workflows into publishable, monetizable apps, starting from any selection on the canvas.

  • A graph planner derives each app’s entry inputs and final outputs from dependency and reachability rules.
  • One-click “Build App” from any canvas selection, with a live preview on hover.
  • Canvas rendering about 20% faster: image, video, and import nodes load only the resolution their zoom needs.
~20%faster canvas rendering
From zoom-aware image and video loading and identity-based memoization.
120 fpspan and zoom on an image- and video-heavy workflow
Measured on the live canvas on a 120 Hz display: median frame 8.3 ms, p95 9.3 ms once media is cached. A team result that includes my rendering work.
Read the case study: App Builder

Generative Fill and Compositor

ImagineArtSenior Frontend EngineerFeb 2026 – Sep 2026

Two AI image editors inside ImagineArt: inpainting and outpainting that respond before the model does, and multi-layer composition on Konva.

  • Inpainting and outpainting with optimistic previews and reusable masks.
  • Konva multi-layer editing with smart guides, blend modes, and undo/redo.
  • AI text-layer extraction, so text in an image can become its own layer.
Read the case study: Generative Fill

cocoding.ai: AI application builder

cocoding.aiSoftware Engineer (part-time)Nov 2024 – Feb 2026

A browser IDE and real-time AI chat that turn a prompt into a working full-stack app, built from the first commit to production.

  • Browser IDE: Monaco, a file explorer, Output and Terminal xterm.js sessions on each chat’s Kubernetes pod, and a live preview.
  • Agent output streamed over SSE into a live timeline, while the preview animates the step in progress.
  • Web performance improved by 80%, reaching Lighthouse 100 in SEO, Accessibility, and Best Practices, and 91 in Performance.
80%web performance improvement
Code splitting, dynamic imports, critical-CSS inlining, and re-render profiling.
100Lighthouse SEO, Accessibility, and Best Practices
Reported scores for the public-facing platform, with 91 Performance.
Read the case study: cocoding.ai

Hashmato: multi-tenant cloud POS

HashmatoSoftware Engineer IISep 2025 – Mar 2026

4payment providers behind one adapter
Stripe, Omise, PayNow QR, and an in-house gateway, chosen per tenant.
30+domain modules on one data-access layer
Database-per-tenant PostgreSQL with TypeORM.
1 queryper page for price overrides
Replaced an N+1 lookup in the pricing engine’s first version.

Payment, delivery, pricing, and ordering services for a database-per-tenant POS, where every tenant brings its own providers.

  • One adapter interface for Stripe, Omise, PayNow QR, and an in-house gateway.
  • Deliveroo, Grab, and Delivery Hero behind one partner interface.
  • A shared data-access layer behind 30+ domain modules on database-per-tenant PostgreSQL.
Read the case study: Hashmato

iGOT.ai: AI agents and workflow platform

iGOT.aiFrontend Engineer, then Engineering ManagerMar 2024 – Mar 2025

The graph engine behind a no-code AI agent builder, then leading the team that built five AI and Web3 products on top of it.

I built

  • Architected the graph engine behind iGOT.ai Studio’s no-code AI agent builder in React Flow, syncing canvas edges with backend execution order.
  • Added auto-layout and optimistic node creation; the engine powers 23 task-node types.

I led

  • Managed and mentored 7 engineers across 5 AI and Web3 products; standardized containerized builds, CI/CD, and pre-commit quality gates.
  • Architected and shipped a full-stack AI audit platform for a Vietnamese government client: a 6-stage multi-agent RAG pipeline, role-based access, and versioned AI results with a full audit trail.

Engineering impact

A number is only useful next to what it measured, so each one keeps its context and its source.

~20%faster canvas rendering
Zoom-aware loading across 6 resolution tiers for image and video nodes, plus identity-based memoization of node renders. The live canvas pans and zooms at 120 fps on a 120 Hz display (median frame 8.3 ms, measured).
ImagineArt Workflows
80%web performance improvement
Code splitting, dynamic imports, critical-CSS inlining, and re-render profiling. Lighthouse: 100 SEO, 100 Accessibility, 100 Best Practices, 91 Performance.
cocoding.ai
100M+users on the platform I shipped into
ImagineArt’s reach across web and mobile. A measure of scale, not of users I acquired.
ImagineArt
2,152commits to ImagineArt’s web app in 7 months
Every commit across all branches, Feb to Sep 2026. The most of 33 contributors in that period.
ImagineArt
7engineers managed and mentored
Across 5 AI and Web3 products, with shared CI/CD and quality gates.
iGOT.ai
16languages
The platform and its MDX blog, localized with next-intl, JSON-LD, and hreflang.
cocoding.ai

Capabilities

Four connected groups. Each one is traced to the projects that actually used it.

  1. Interactive systems

    Node-based editors, layered canvases, and graph logic that stay fast when the canvas gets crowded.

    React, React Flow, Konva, HTML Canvas, ELK.js, Graph algorithms, Rendering performance

    Used in

  2. AI products

    Creative tools, agent builders, and app generators where a model does the work and the interface makes it legible.

    LLM integration, Multi-agent workflows, RAG, Streaming interfaces, AI image and video tools, CrewAI

    Used in

  3. Full stack engineering

    Typed products end to end, from the React tree to the API, the queue, and the database.

    Next.js, TypeScript, Node.js, NestJS, Python, FastAPI, PostgreSQL, MongoDB, Redis, WebSockets, SSE

    Used in

  4. Platform and architecture

    Multi-tenant SaaS, payments, partner integrations, and the delivery pipeline that ships them.

    Multi-tenant SaaS, Payments, Integrations, Docker, Kubernetes, Helm, AWS, GitHub Actions, System design

    Used in

About

I’m a senior full stack engineer in Dhaka, Bangladesh. For more than seven years I’ve worked on products where the difficult part hides behind the interface.

Workflow editors with dozens of node types. AI tools that have to respond while a model is still working. Multi-tenant backends where every tenant brings its own payment provider. I like these problems because the frontend and the backend can’t be solved separately.

I take features from the first idea to production: rendering performance in the browser, the API behind it, the database, and the pipeline that ships it. I’ve done that remotely with teams in the US, Singapore, France, and Vietnam, and as an engineering manager leading seven engineers at iGOT.ai.

More about how I work
Remote teams I’ve shipped with, from Dhaka: United States (ImagineArt, BeWeddy), Singapore (Hashmato), France (cocoding.ai), Vietnam (iGOT.ai).