Name the use case.
The workflow gets built.
cases built

AI Automation Engineer & ComfyUI Expert
Practical AI systems & automations for small and growing businesses — simplifying operations, winning more customers, and scaling what works.
AI that actually moves the needle.
With over 1.5 years building around AI, I'm focused on helping SMBs evolve with this technology.
As someone deeply into AI, coding, and real-world tech solutions, I see this shift as a massive opportunity. My goal is to help businesses adapt — using AI to simplify operations, improve customer experiences, and build stronger connections between brands and people. It's not about tech for its own sake; it's about practical, profitable, problem-solving solutions that help businesses grow smarter.
Now I'm channeling everything I've learned into one mission: helping SMBs evolve with AI. I build what actually moves the needle.
50+
Projects
AI
Focus
10+
Tools
Projects Completed
Hours Saved
AI Models Trained
Years Experience
ComfyUI Pipelines
Flux, SDXL, Wan Models
EXPERT
AI Automation
Agentic Workflows & Integration
ADVANCED
AI Model Training
Custom LoRA, Fine-tuning, Voice Cloning
SPECIALIST
Hyper-real people and lifestyle imagery — and this is just a taste. 100% AI-generated.
22+ lifestyle visuals — just a preview
300+ more across interiors, products, jewellery, weddings & beyond
10 categories · art, interiors, materials, staging, fashion, jewellery, wedding & more
AI-powered tools deployed in production — replacing entire workflows, teams, and timelines.
I design it, I build it, I ship it. AI systems, automations, brands, visuals... if it can be made, I'm probably already making it.
I’m obsessed with building. I bounce between AI systems, automations, and visuals, and I’m pretty much always working on something.
AI made making easy. Taste is what's rare now. The same eye behind these goes into every system, dashboard, and brand I ship.
Open source at the centre. ComfyUI is open-source software, and everything here is built and optimised for open-source models, so most of it runs on your own hardware with no per-call bill. A few pieces use paid services where they earn it. The same systems can be rebuilt with custom scripts, n8n or Make.com: the tool is a choice, not a constraint.
Custom ComfyUI node graphs running in production, the automation systems built around them, and the finished films they produce. The systems are not graphs of their own: they are powered by n8n + comfyui + ai. Tap any tile for the full breakdown and clips.
No demo clip yet
18 graphs · 9 systems · 9 output pieces · 27 playing inline · 15 run autonomously
ComfyUI workflows and AI image workflows across different models, all of them starting from material the client already has — a phone photo, a reference they liked, a moodboard, a floor plan. No reshoot and no render house. The five below are examples of the kinds of input that work, not the list of what is possible.













Any subject, on any of these 18 base models. A LoRA is trained from scratch; a finetune adapts existing weights. Every model listed supports both.
Create new images & videos from text or reference inputs
Modify, enhance, or transform existing images with instruction-based editing
Photorealistic outputs trained on real-world data for lifelike results
Every pair below is a model I trained: the left half is what went in, the right half is what came out. Grab the handle and wipe across.


Removes furniture from interior shots for staging


Transforms photos into exaggerated caricature artwork


Generates styled animal portraits from reference photos


Removes furniture from interior shots for staging


Transforms photos into exaggerated caricature artwork


Generates styled animal portraits from reference photos
ComfyUI custom node development in Python. Open source, and in use across the ComfyUI community.
Open source, and in use across the ComfyUI community.
19 custom nodes written in Python.
I build autonomous AI employees, each with their own phone number, own WhatsApp, and own voice. They cold-call leads, send WhatsApp outreach, and do voice calls, all with unique personalities. Watch them on a live dashboard, assign tasks, see the org chart.
They don't wait for messages. They reach out.
Powered by OpenClaw or Hermes Agent: open-source, self-hosted AI agents with tool calling, MCP integrations, and 5,400+ installable skills
Click any capability to learn more
REAL PHONE CALLS
The main admin agent and approved agents get their own Twilio phone numbers and ElevenLabs voices. They cold-call leads, answer inbound calls, and sound like real people, each with a unique, cloned or designed voice.
One agent instance, every messaging platform simultaneously
Click any screen to expand
Install any skill and the agent becomes an expert at it instantly, from web scraping to smart home control
The Architecture
Everything is location-agnostic. ComfyUI on your local GPU, RunPod, Replicate, or any API endpoint: the agent triggers it from anywhere. n8n on a VPS, local machine, or cloud: the agent calls any workflow via API endpoint, webhook, or direct connection. The agent itself can be hosted anywhere too. It can even spin up its own endpoints when needed. All three (agent, ComfyUI, n8n) can be on completely different servers and still work together seamlessly.
Not demos. Not notebooks. Actual serverless endpoints, auto-scaling GPU clusters, and Cog-packaged models running in production.
Any workflow, any complexity, exposed as a REST endpoint
SDXL, Flux, Wan, Z-Image: trained and deployed for production
Full preprocessing → inference → post-processing chains
Depth, canny, pose, segmentation: pick per use case, deploy as API
Auto-scaling serverless inference, scale-to-zero
Package any model into a Cog container, deploy as API
Deploy full ComfyUI workflows as hosted API endpoints
Cutting-edge technologies powering next-generation AI solutions
ComfyUI
n8n
Python
JavaScript
OpenAI
Claude
ComfyUI
n8n
Python
JavaScript
OpenAI
Claude
ComfyUI
n8n
Python
JavaScript
OpenAI
Claude
Google Cloud
Replicate
Kling AI
GitHub
Android Studio
Antigravity
Google Cloud
Replicate
Kling AI
GitHub
Android Studio
Antigravity
Google Cloud
Replicate
Kling AI
GitHub
Android Studio
Antigravity
Let's turn your wildest creative ideas into reality with AI that actually works
LET'S COLLABORATE“Great knowledge of automation, hard worker, went above expectation”
— Benjamin
“I am a doctor, and I post content for medical students. I was posting blindly, without really thinking about it. Hisham optimized my content creation process with AI, and changed my perspective on how to think about content. Once I started going in the direction he pointed me, the second reel pulled views and engagement I had never had before. Since then every reel lands, and a couple of months on, around 100k views has become my baseline. I now have enough confidence to do it myself — which is what I actually needed.”
— Imran
“Great knowledge of AI, what you built helped me bring in more customers and serve them better”
— Lokesh
“Working with you was a hassle-free experience. Loved the way you solved our problem and helped us build an organized system around it. We are not worried about things messing up.”
— Robert
“We loved the batch processing automation you built — it was super accurate on exactly what we wanted for our products. Impressive that you pulled it off with local ComfyUI models, and it saved us a lot of budget.”
— Dave
Schedule a 20-minute discovery call to discuss your project. Let's turn your wildest creative ideas into reality with AI that actually works.
Free discovery call to discuss your project needs
Video conference info added after booking
Get instant confirmation and meeting details
Available slots updated in real-time
The Archive
Earlier AI experiments — preserved as the field keeps evolving.
WHAT'S UP
Join the community and stay updated with the latest AI automation content
LINKEDIN
@hishamahmer
INSTAGRAM
@hishamahmerrr
YOUTUBE
@thehishamahmer