Skip to content

Welcome to the

0
Azan Akram in a charcoal blazer over a black t-shirt

Azan Akram

Agentic AI Engineer & Solution Architect

Azan AkramAgentic AI Engineer & Solution Architect

Digital employees, not chatbots.

Now buildingmulti-agent systems

I design multi-agent systems, LLM products and the full stack around them.

  • 10+production projects shipped
  • 3+years of engineering
  • 2multi-agent platforms architected

I take AI products from whiteboard to production. I am an Agentic AI Engineer & Solution Architect. I design, architect and ship end-to-end AI platforms and multi-agent systems that hold up in production.

Who I am

Right now I architect a platform that provisions digital employees: autonomous AI workers with tools, permissions and accountability. I also designed a multi-agent assistant that runs the whole job search for its users.

Before the agentic era I earned my foundations the traditional way: a finance platform for a client in the Netherlands, an AI-powered e-commerce store, and real-time coding and banking platforms during my internship.

Outside client work I explore new LLM tooling, contribute to AI engineering communities and mentor developers moving into agentic AI.

Based in
Faisalabad, Pakistan
Works with
Founders and teams, worldwide
Focus
Multi-agent systems, LLM products
Currently
Architecting digital employees
How a build runs: your idea, your data and your tools go to Azan, who plans the work with four agents, a planner, a researcher, a builder and a reviewer. The reviewer sends one fix back to the builder, then approves, and the result ships as agents, APIs and a product.

Drag to spin. Point at a name to bring its logo forward.

What I build with

Languages

  • Python
  • TypeScript
  • JavaScript
  • PHP

AI

  • Claude
  • Anthropic
  • OpenAI SDK
  • Gemini
  • LangChain
  • LangGraph
  • Hugging Face
  • ElevenLabs
  • Deepgram
  • n8n
  • ChromaDB

Backend

  • FastAPI
  • Django
  • Flask
  • Node.js
  • Laravel
  • Livewire

Frontend

  • React
  • Next.js
  • Tailwind CSS
  • Vite

Data

  • PostgreSQL
  • MySQL
  • Redis
  • Supabase

Cloud

  • Docker
  • Vercel
  • Google Cloud

Tools

  • Git
  • GitHub
  • Cursor
  • Postman

Agentic systems in production

Agentic work

Three systems, from the orchestration layer down to the tools. Select any node to see what it does.

01 / 032025Solution architectConfidential client work

Crewline

A platform that provisions digital employees.

The problem. Businesses wanted AI that finishes tasks inside their own tools, not another chat window.

  • Architected the multi-agent orchestration layer, the agent runtime and the platform connectors.
  • Designed a tool and skill registry so new agents deploy fast, with access control and observability built in.
  • The core idea: not a chatbot, a provisioned digital worker with tools, permissions and accountability.
  • Python
  • FastAPI
  • OpenAI SDK
  • Multi-agent systems
  • Next.js
  • PostgreSQL
Business requestOrchestratorPlannerResearcherExecutorReviewerSkill registryCRMEmailDocsAccess controlTracing

Select any node to see what it does.

Illustrative run, simplified
  1. 00.00orchestratortask received: reconcile March invoices
  2. 00.41plannerplan: fetch ledger, match invoices, flag gaps
  3. 02.10executortool call: erp.fetch_ledger(month=3)
  4. 04.52reviewer3 mismatches flagged for a human

02 / 032025Solution architectConfidential client work

Pathfinder

A multi-agent assistant for the whole job search.

The problem. Job seekers repeat the same research, tailoring and follow-up for every single application.

  • Specialist agents for resume tailoring, job matching, application drafting and interview coaching.
  • An orchestrator coordinates them, so each step hands clean context to the next.
  • Personalised retrieval grounds every application in the candidate's real history.
  • Python
  • FastAPI
  • OpenAI SDK
  • Agentic workflows
  • Next.js
  • PostgreSQL
Candidate historyOrchestratorResume optimizerJob matcherApplication writerInterview coachRetrieval storeJob boardsEmailCalendar

Select any node to see what it does.

Illustrative run, simplified
  1. 00.00orchestratornew target role: senior backend engineer
  2. 00.18matcher12 openings ranked by fit
  3. 01.05optimizerresume tailored for the top 3
  4. 01.47writerapplications drafted, waiting for approval

03 / 032025Designer and developerPIAIC project at NTU

Study Mode

Gemini-powered study sessions built from your own notes.

The problem. Students collect notes but rarely turn them into active practice.

  • Integrated the Gemini API to personalise note-taking, quizzes and task planning.
  • Automated study sessions with n8n workflows that explain, quiz and review.
  • Engagement on the beta tripled in the first two weeks.
  • Gemini API
  • n8n
  • Tailwind CSS
  • HTML
Study Mode screenshot 1Study Mode screenshot 2
Open the live app
Student notesGeminiExplainerQuiz makerPlannern8n workflowsFlashcardsTasks

Select any node to see what it does.

Experiments

Labs

Working ideas I build to test new agent patterns. Open any of them to watch a simulated run.

Where the engineering
came from

6 full-stack builds from before the agentic era. All of them on one page

2025Client, Sahiwal

ZapBoots

E-commerce platform with AI product recommendations and smart inventory.

Laravel / Livewire / MySQL

Visit the live site
01 / 06

Experience

Where I have worked

2025 TO NOWAgentic AI Engineer & Solution Architect

AI product company Under NDA

  1. 2022 to 2023

    Backend and frontend intern

    Webi Solutions, Internship

    Built CodeArena, ShareFlow and BankPro, and learned clean architecture the hard way.

    • WebSocket live judging
    • Secure access control
    • MFA and fraud detection

    Laravel / Tailwind CSS / MySQL

  2. 2023 to 2025

    Full-stack developer

    Freelance and client work, Netherlands and Pakistan

    Shipped FinTrack, ZapBoots, FOOPA and Study Mode for clients and for myself.

    • AI features inside real products
    • Laravel and Livewire backends
    • Responsive, fast frontends

    Laravel / Livewire / MySQL / Gemini API

  3. 2025 to now

    Agentic AI Engineer & Solution Architect

    AI product company, Under NDA

    Architecting a platform that provisions digital employees, plus a multi-agent career assistant.

    • Multi-agent orchestration and agent runtime
    • Tool and skill registry with access control
    • Observability from day one

    Python / FastAPI / OpenAI SDK / Next.js / PostgreSQL

Education

The path that led here, one milestone at a time.

  1. CAST High School, KIPS Coaching

    Matriculation in Computer Science

    1123 out of 1200. Best Thesis Awardee and Emerging Student Award.

  2. Government College University Faisalabad

    Intermediate in Computer Science

    Computer science track at GCUF.

  3. PIAIC

    Agentic AI program

    Agent Factory fundamentals, prompt and context engineering.

  4. Today

    Agentic AI Engineer & Solution Architect

    Designing digital employees for real businesses.

Certificates

18 certificates, 8 of them from Anthropic's Claude Academy. See the full list

Services

Fixed scope, clear price, one person accountable from architecture to launch. Pick a tier, or just tell AzanGPT what you need.

  1. 01

    Tell AzanGPT

    A few minutes of chat, no form

  2. 02

    Scope and estimate

    What is practical, and a price range

  3. 03

    Book a weekend call

    30 minutes on Google Meet

  4. 04

    Fixed quote

    One price, agreed before work starts

  5. 05

    Build with weekly demos

    Working software every week

  6. 06

    Launch and support

    Deployment, monitoring and evals

Scope your project with AzanGPT

Describe the project in plain words, and get an honest scope and a price range.

Online

Hi, I'm AzanGPT. Tell me what you want to build and who it is for. I'll say honestly what is practical, give you an estimate from Azan's own price formula, and if it fits, book a 30 minute call with him.

Enter sends the message; Shift and Enter start a new line.

Prefer a form? , or email Azan.

What clients say

01 / 03

Azan built AI-driven insights into FinTrack that cut manual categorisation by over 80%. The dashboards are crisp, fast and genuinely useful.

Aisha KhanProduct Lead, FinTrack

Let's build something intelligent.

A multi-agent platform, an LLM inside your product or a full build. The first consultation is free. Tell AzanGPT about it, and it books the call with Azan.

I reply to every message within 24 hours.

Tell AzanGPT

It scopes the project, checks Azan's calendar and books your call.

Online

Hi, I'm AzanGPT. Tell me what you want to build and who it is for. I'll say honestly what is practical, give you an estimate from Azan's own price formula, and if it fits, book a 30 minute call with him.

Scope summaryDraft
Goal
Waiting
Who uses it
Waiting
Connects to
Waiting
What is practical
Waiting
Service and estimate
Waiting
Budget stated
Waiting
Next step
Waiting
Enter sends the message; Shift and Enter start a new line.

Prefer a form? , or email Azan.

visitor@azan-portfolio: ~

_ ___ ___ _____

/_\ ____ __ _ _ _ / __| _ \_ _|

/ _ \ |_ / / _` || ' \ | (_ | _/ | |

/_/ \_\/__| \__,_||_||_| \___|_| |_|

Hi, I am AzanGPT. Ask me about Azan's work, stack, services or how to reach him.

Type 'help' for commands, or ask a question. Tab completes, the arrow keys walk your history.

AzanGPT runs on an AI model that knows this site. It can get things wrong, so check anything important with Azan.