Home / Projects / VideoGen AI
In Development · AI Video Generation

VideoGen AI: text & image to video on an RTX 4090

Ahmad Humayoun is building VideoGen AI — a platform that generates video from text prompts or images using the Wan 2.2 model, orchestrated on a RunPod RTX 4090 with a production queue. The architecture runs on Next.js, FastAPI, Redis and Celery — the same disciplined stack that powers his shipped products.

Wan 2.2state-of-the-art video model
RTX 4090GPU on RunPod
Text → Videoand Image → Video
Queue-basedCelery + Redis
Honest status: this page documents a project in active development. It is published to show how Ahmad works — architecture first, shipped later — and will be updated when the platform launches.

Why video generation, why now

Video is the most expensive content to produce, and AI video models are now good enough to automate huge parts of it — ads, product demos, social content, training material. Most teams that want AI video are blocked by two things: knowing how to run the models, and running them reliably at scale. This project tackles both.

The build

[Screenshot — generation preview, in development]

What "in development" means here

Clients often ask whether Ahmad can build video AI. This page is the honest answer: the model runs, the pipeline is built, and the product is being packaged. It is published deliberately — the same way this portfolio shows real numbers everywhere else, it also shows real status. What's proven today: model integration, GPU orchestration and queue infrastructure.

What this means for you

If your team needs AI video — ads at scale, automated product demos, training content — the pipeline being built here is directly reusable for your project. Ahmad ships AI into production rather than presenting slideware, and this project is the latest proof.

Need AI video, or AI product engineering?

Book a call to discuss what AI video could do for your content or product.

Book a Free Consultation WhatsApp +92 303 4271843