Software Engineering, AI & Automation

Software, AI and automation
for businesses.

We build websites and apps, train models and automate repetitive work.

Software Engineering · AI Engineering · Automation & Integrations

Independent. Direct technical accountability.Task to evaluation. Defined scope.Based in Malaysia. Working internationally.

Focus this row to pause scrolling. Moving focus away resumes it.

  • Astro
  • Next.js
  • Go
  • Python
  • PyTorch
  • Hugging Face
  • scikit-learn
  • n8n
  • OpenRouter
  • Docker
  • Kubernetes
  • AWS
  • Google Cloud

Where we can help

A clear idea.
A system that works.

Connect what people need to do with the software, data and infrastructure needed to support it.

  • A website or application that needs a defined build, dependable backend and a clear path to launch.
  • A prediction or classification task that needs a measurable baseline and better results.
  • A language-model task that needs suitable data, evaluation and a defined operating boundary.
  • Business information repeatedly copied and checked between tools.
  • Documents and external data that need structured extraction and review.
  • AI and automation processes whose failures need to be visible and recoverable.

Three focused pillars

Build the software.
Connect the work.

Start with a software project, a model task or a repeated process. We agree on useful outputs, evaluation criteria and the responsibilities around delivery.

01

Software Engineering

The product needs to work beyond the screens, with dependable data, integrations and a clear path to deployment.

Build websites, web applications and the backend services they need, from agreed requirements and approved designs.

  • Web applications and operational systems
  • Backend services, APIs and integrations
  • Responsive interfaces, stored data and permissions
  • Testing, deployment, monitoring and handover

Scoped quotation

Build and launch your system
02

AI Engineering

You have a defined task and need a model whose results can be measured against it.

Develop and evaluate models for prediction, classification and language tasks, with the data, compute and deployment requirements agreed upfront.

  • Prediction and classification models
  • Language-model training and fine-tuning
  • Dataset preparation, baselines and evaluation
  • Inference, operational checks and handover

Scoped quotation

Discuss AI engineering
03

Automation & Integrations

Information gets copied, checked and passed between tools by hand.

Connect existing tools and automate a repeatable process, with useful outputs, visible failures and clear human handoffs.

  • Workflow automation and tool connections
  • API integrations and data synchronization
  • AI-assisted scraping and structured extraction
  • Scheduled reporting, notifications and follow-up

Scoped quotation

Discuss an automation
  1. 01

    Define the scope

    Identify the users, task and desired outputs. Agree on deliverables and observable acceptance criteria.

  2. 02

    Plan the implementation

    Review requirements, supplied designs, data and integration access. Agree on the implementation, evaluation and delivery plan.

  3. 03

    Build & evaluate

    Build the agreed software, model or workflow. Check the user journeys, results, failure paths and human decisions.

  4. 04

    Deploy & hand over

    Release into the agreed environment, verify the critical paths and document operational ownership and recovery.

Selected engineering work

Real problems.
Working software.

Explore all work
Pastelocity e-commerce website shown on a desktop display

E-commerce operations

Pastelocity

Connected customer, order, payment, receipt and administrative workflows in one business operations system.

Read the case study
Memora website presented on a desktop display

Event technology

Memora

A private QR-based guest contribution experience, with media handling, event access and export workflows.

Read the case study

Engineering Lab

The infrastructure
behind useful AI.

The engineering work examines how software executes, preserves state, exposes failures and recovers. Forge documents a working private AI system; distributed systems and AI workload infrastructure guide further investigation.

Explore systems & AI infrastructure
Axelyn ForgeWorking private system

AI proposes.
The system verifies.

  1. 01
    Verified evidenceCandidate facts & sources
  2. 02
    Model proposalConstrained rewrite operations
  3. 03
    ValidationDeterministic checks & rendering
  4. 04
    PublicationComplete output package
Inspect the published architecture
Amin Haiqal, software engineer behind AxelynAmin Haiqal / Software engineer

The engineer behind the work

Directly accountable engineering.

I'm Amin Haiqal, the software engineer behind Axelyn. My experience spans backend services, data pipelines, API integration, production systems and applied AI. My technical focus is the infrastructure and reliability behind those systems.

I work directly with you, from defining the implementation to deployment and handover. Based in Malaysia, I work with suitable clients internationally.

Explore my portfolio