I build AI-powered systems that actually run.
I turn business processes and ideas into working automations, internal tools and web systems — using AI-assisted development, real APIs and practical software engineering.
Python · Node.js · APIs · n8n · Claude · Docker · WordPress
Systems I’ve built and shipped
Five case studies spanning client delivery, production systems running today, and AI-orchestrated automation. Sensitive project details are sanitized or diagrammed — see each case study for what that means.
Autonomous Content Pipeline
Idea → published video, end to end
A fully automated content pipeline that takes a topic from research to a scheduled, published video with no manual editing — script generation, voice, imagery, assembly and publishing all orchestrated in one system.
Multi-Tenant Fund Management Platform
Real clients, running in production
A business management platform for community lending/savings cooperatives — member records, approval workflows, automated WhatsApp and email reminders, and reporting. Two independent client deployments run today.
Almagzal — Arabic Education Website
Custom WordPress build, delivered to a paying client
A custom WordPress site for an Arabic-language education brand — custom theme and content-type work, full RTL implementation, and a shipped, controlled production deployment.
Approval-Gated Social Content Automation
AI drafts it. A human still approves it.
A workflow that turns structured content into rendered social carousels, routes them through Slack for human approval, and only then publishes — automation with a deliberate checkpoint before anything goes live.
AI-Powered Operations Dashboard
One screen, and a briefing written by Claude
A status dashboard that pulls health and metrics from several independent systems into one view, then runs a scheduled, guard-railed Claude Code CLI job every morning to write a plain-English operations briefing.
Selected Technical Work
Smaller technical experiments, included as evidence of engineering range rather than as headline projects.
Applied ML — Strategy Orchestration Engine
A multi-strategy trading system with a custom reinforcement-learning training loop (PPO/SAC) using population-based training to tune its own policy over time, backed by live data pipelines and a risk-management layer. Shown here purely as ML/engineering evidence — no performance figures or live data are shared, and this is not offered as a trading or investment service.
Arabic AI Video Processing
A pipeline that transcribes casually-spoken Gulf Arabic, rewrites it into formal Modern Standard Arabic, and burns karaoke-style right-to-left captions onto vertical video — solving real ASR and bidirectional-text rendering problems that most video tooling handles poorly.
Safe-by-Design Momentum Scanner
A market scanner deliberately built to never hold exchange credentials or place trades — advice-only by design, with real automated test coverage and a containerized deployment. Included as evidence of engineering discipline, not as a trading tool.
Capabilities
AI & Automation
LLM-driven workflows, AI agents, content pipelines and human-in-the-loop systems that take a defined process off a person’s plate.
Internal Tools
Dashboards, business systems and operational tooling that replace spreadsheets and manual checklists with something that runs itself.
Integrations
APIs, webhooks and third-party services wired together into scheduled, reliable processes — exchanges, messaging platforms, CRMs, analytics.
Web Implementation
Custom web applications, backend systems and WordPress implementation — built, tested and deployed, not just designed.
AI-assisted, human-directed.
AI-assisted development is part of my workflow, not the product. I define the problem, direct the implementation, inspect results, test, debug, refine and deploy — and take responsibility for the outcome. The product is a working system that solves the problem.
Understand
Turn the business problem into a clear system requirement — inputs, outputs, and what "working" actually means.
Build
Use AI-assisted development, code, APIs and automation tools to implement quickly, directing every step of the build.
Verify
Test, debug and inspect the actual behavior of the system — never trust generated code blindly.
Ship
Deploy a working implementation, then iterate based on how it actually gets used.
Between the business problem and the implementation
I’m Ali, a Computer Engineering graduate based in Kuwait. I build practical AI and automation systems — content pipelines, internal tools, business software, API integrations — and I use AI coding agents like Claude Code as leverage to move from idea to working implementation quickly.
My work sits between the business problem and the code. I use AI-assisted development to build fast, then I test, debug and ship the resulting systems myself — across Python, Node.js, APIs, automation tooling and web platforms including WordPress.
Have a process that should be automated?
I’m available for contract work, remote projects, agency overflow and white-label implementation — AI automation, internal tools, integrations and web systems.