Stop prompting.
Start engineering.
A 30-hour professional development course that transforms developers into disciplined AI-assisted engineers — from architecture through production deployment using Claude Code.
Most developers are
using AI wrong.
- Copy-pasting from chat and hoping it works
- No CLAUDE.md — Claude Code guesses your architecture on every session
- Skipping tests because "it looks right"
- Deploying with a prayer instead of a pipeline
"This course teaches the discipline that makes AI a precision instrument — not a lucky guess machine."
Before this course → After this course
BEFORE
- Prompting and hoping
- Rewriting the same context every session
- Code that works until it doesn't
- Manual deployments that break
- Zero test coverage
- Security as an afterthought
AFTER
- Engineering with a disciplined AI workflow
- CLAUDE.md rules that persist across every session
- TDD: tests written before code, verified by a machine
- Push-to-deploy CI/CD pipeline in GitHub Actions
- Three-layer test suite (unit + integration + E2E)
- OWASP Top 10 security hardening — automated
The Curriculum
10 sessions × 3 hours each. Theory + live demo + hands-on exercise in every session.
⏱ 3 hours
- Installation, configuration, and IDE integration for Claude Code
- Mental model: how Claude Code reads context, tools, and project structure
- Writing your first CLAUDE.md — the most impactful file you'll create
- Exercise: Configure a CLAUDE.md for your own project
⏱ 3 hours
- Writing PRDs and architecture docs that Claude Code can execute against
- System design with Claude: ERDs, API contracts, component boundaries
- Decision logging: capturing architectural choices in your project
- Exercise: Generate a PRD and architecture diagram for TaskFlow
⏱ 3 hours
- Schema design: normalization, indexing strategy, multi-tenancy patterns
- Using Claude Code to generate and review migrations
- Query optimization: explaining plans, identifying N+1 problems
- Exercise: Design and migrate the TaskFlow PostgreSQL schema
⏱ 3 hours
- Test-driven development: red-green-refactor with Claude Code
- API design: REST conventions, error handling, validation patterns
- Integration tests: testing database operations against real data
- Exercise: Build and test TaskFlow API endpoints with 80%+ coverage
⏱ 3 hours
- Component-driven UI development with Claude Code
- Design systems in code: tokens, patterns, enforced consistency
- Frontend TDD: component testing and interaction testing
- Exercise: Build the TaskFlow React UI with a consistent design system
⏱ 3 hours
- OWASP Top 10: identifying and patching the most critical vulnerabilities
- Using Claude Code for security audits: what to ask, how to verify
- Authentication, authorization, and secrets management patterns
- Exercise: Full security audit of TaskFlow, patch all findings
⏱ 3 hours
- E2E testing philosophy: what to test, what to skip
- Playwright setup, writing and maintaining browser tests
- CI integration: tests that run automatically on every push
- Exercise: Write E2E tests for TaskFlow's critical user paths
⏱ 3 hours
- Docker fundamentals: images, containers, multi-stage builds
- Writing production Dockerfiles: security and size optimization
- Docker Compose for local development: databases, services, networking
- Exercise: Containerize TaskFlow with a multi-stage Dockerfile under 100MB
⏱ 3 hours
- VPS provisioning: server setup, SSH hardening, firewall rules
- nginx as reverse proxy: HTTPS with Let's Encrypt, headers, caching
- GitHub Actions CI/CD: automated test, build, and deploy pipeline
- Exercise: Deploy TaskFlow to a live HTTPS URL with one-push deployment
⏱ 3 hours
- Production monitoring: logs, metrics, alerting, and on-call discipline
- Scaling patterns: horizontal scaling, caching, queue-based architecture
- Post-mortems, runbooks, and the production discipline mindset
- Exercise: Set up monitoring and write a runbook for TaskFlow
You ship one real app across all 10 modules.
TaskFlow — a multi-tenant project management SaaS — is built from PRD to production over the 5-week course. By the end, you have a live HTTPS URL, a CI/CD pipeline, and 47+ tests passing.
HTTPS Live URL
Your app is on the internet by session 9
CI/CD Pipeline
One git push deploys to production
47 Tests Passing
Three test layers, all green
Is this course right for you?
This is for you if
- You are a working developer (any level, junior to senior)
- You have used AI coding tools but feel your workflow is inconsistent
- You want to ship production-grade code faster without lowering standards
- You understand basic git, JavaScript/Python concepts, and command line
- You want a deployable portfolio project at the end
Not for you if
- You have never written code before (this is not an intro to programming)
- You are looking for a course on AI theory or LLM internals
- You want a quick certification with no hands-on work
- You expect to watch passively — every session requires active participation
About George E. Salloum
George E. Salloum
Founder, I-Stay SAL
Course Instructor
George has spent 30+ years building and shipping production systems across the MENA region. He designed this course after observing a consistent pattern: developers using AI tools without the engineering discipline to verify, secure, and deploy what they build.
This course is the curriculum he wished existed when AI coding tools first appeared. Every module is built on real production experience — not theory.
Invest in the discipline.
Ship better software.
FULL COURSE
$1,297
per participant
Pricing reflects your local market
- 30 hours live instruction
- 10 session decks (PDF)
- Participant workbook
- 47 asset files (templates, scripts)
- Course GitHub repository access
- Post-course Q&A access (30 days)
- Certificate of completion
Pricing reflects your local market. Team discount available for 3+ seats — email george.s@frontal.ltd
Frequently Asked Questions
Mixed levels. Junior developers learn structured engineering discipline. Senior developers learn to delegate routine work and focus on architectural decisions. Every exercise has a junior path and a senior path.
No. Module 01 covers installation and all foundational concepts. If you have used Claude Code before, you will skip ahead quickly in the first session.
The course is stack-agnostic — Claude Code selects the stack based on your project's requirements. The demo app (TaskFlow) uses Node.js + React + PostgreSQL, but participants can work with any stack they choose for their personal mini-app.
A computer with Git and VS Code installed. A Claude Pro or Max account (approximately $20/month). A basic understanding of JavaScript or Python. All other tools are set up during Module 01.
3 hours per live session, twice a week (5 weeks total = 30 hours). Each session has a small follow-up exercise that takes 30–60 minutes to complete before the next session.
Yes. All live sessions are recorded and available to enrolled participants for 90 days after the course ends.
You can watch the recording. The exercises are designed to be completed asynchronously. A cohort messaging channel keeps you connected.
Full refund within 7 days of the course start date if you have attended fewer than 2 sessions.
The gap between a developer who uses AI
and one who engineers with it
is this course wide.
Next cohort: [DATE]. [N] seats remaining.
Questions? Email george.s@frontal.ltd