Hesham
ServicesWorkVenturesAbout
Ask Hesham
CASE STUDIES · REAL PRODUCTION BUILDS

Impossible workflows, documented end-to-end.

Every build below is a real production system — with real node counts, real outcomes, and a full technical write-up to back it up. This is the complete index.

Case study

An AI Virtual Accounting Department for Financial Document Processing

A 77-node n8n workflow that uses Gemini OCR to extract invoices, receipts, and bank statements into Google Sheets — with auto-categorization and a natural-language financial-analysis agent.

Read the case study →
Case study

Digital Art to Shopify: AI-Generated Product Listings from Images

A 34-node n8n workflow that turns digital art images into SEO-optimized Shopify listings — AI vision analysis plus LLM-generated content — tracked end-to-end in Airtable.

Read the case study →
Case study

A Daily AI Social Content Engine for a Beauty Salon

A 63-node n8n engine that writes daily on-brand beauty-salon posts with matching AI images and auto-publishes them across Telegram, WordPress, X, Facebook, and LinkedIn.

Read the case study →
Case study

A Modular Viral Content Engine for LinkedIn & X with Human Approval

A 156-node n8n engine that discovers viral ideas, generates on-brand LinkedIn and X posts with AI images, gates them through Telegram approval, and auto-publishes to both platforms.

Read the case study →
Case study

A Weekly AI Research Trend Alerter Powered by Agentic RAG

A 46-node n8n workflow that weekly fetches arXiv AI papers, classifies them with an LLM, stores them in Weaviate, and uses agentic RAG to email a trend digest.

Read the case study →
Case study

A Scalable Sync Architecture for High-Volume CRM Integrations

A modular, 244-node n8n reference architecture for scalable CRM sync workflows — graded trigger, polling, backfill, and error-handling patterns that prevent rate limits, duplicate runs, and silent failures.

Read the case study →
Case study

A Daily F1 Race-Winner Prediction Engine with Confidence-Gated Alerts

A 38-node n8n pipeline that predicts F1 race winners daily — fusing live data, eight driver metrics, RAG over race history, and GPT-4o — with confidence-gated Slack alerts.

Read the case study →
Case study

An AI Social Media Campaign Engine with Human-in-the-Loop Approval

A 63-node n8n workflow that generates on-brand LinkedIn and Facebook campaigns with AI copy and imagery, gates every image through Slack human approval, and schedules posts via Late.

Read the case study →
Case study

Evergreen Digital Assets: Dynamic Images & Smart URL Redirects

A compact 19-node n8n workflow that keeps distributed images and links evergreen — fixing their addresses while updating their content on a schedule, so published assets never go stale.

Read the case study →
Case study

An Autonomous AI Avatar Engine for Short-Form Investing Video

A 234-node autonomous n8n pipeline that researches, scripts, voices, animates, and publishes a daily AI-presenter investing video to YouTube and TikTok — entirely on self-hosted infrastructure.

Read the case study →
Case study

An Autonomous AI Financial Blogger & Newsletter Engine

A 115-node, fully autonomous n8n pipeline that researches, writes, illustrates, and publishes a daily voice-driven financial blog post to Ghost — entirely on self-hosted infrastructure, with zero human input.

Read the case study →
Your turn

Got a workflow no one’s been able to build?

A free 20-minute call. Bring the messy, ambitious, “is this even possible” version, and I’ll tell you if it’s buildable and roughly how.

Book the call
[ GitHub ]·[ Udemy ]·[ Upwork ]·[ Substack ]·[ YouTube ]·[ LinkedIn ]