About
Turn prompts into ProRes in under 5 minutes
We build AI video tools that respect deadlines, budgets, and the humans doing the work—so creative teams can ship more without burning out.
Unionvora started in Brooklyn, November 2023. Founder Luis Navarro spent eight years as a post-production supervisor at an ad agency, watching editors waste 60% of their week on grunt work—sourcing stock, rendering roughs, resizing for platforms. When Stable Diffusion's video models dropped, Luis saw the gap: every AI tool targeted hobbyists, not production teams with deadlines. He built the first T2V API in four months. Twelve agencies signed up week one.
We run render farms on NVIDIA A100 GPUs in US-East and US-West AWS regions. Every model update passes a 200-clip regression test before deployment. We don't train on scraped social media—our datasets come from licensed stock (Shutterstock, Pexels) and public-domain archives. Our support team includes two former video editors who answer technical questions in under 15 minutes during business hours. We publish our roadmap publicly and let customers vote on features.
Our path
- Nov 2023Luis Navarro launches T2V beta with 12 agency customers in Brooklyn
- Feb 2024I2V and V2V pipelines go live; 500 users rendering 2,400 videos/month
- Jun 2024$3.2M seed round led by Craft Ventures; API launches with 18 launch partners
- Oct 2024Lip-sync and voice cloning added; first Enterprise customer (120-seat agency)
- Jan 20254K upscaling and 60fps interpolation released; render farm scales to 240 GPUs
- May 2025Custom model training (LoRA fine-tuning) available on Enterprise tier
What we stand for
We publish our uptime stats monthly (no cherry-picking) and guarantee render-time SLAs on Enterprise. When we miss a target, we credit your account before you ask.
We don't claim our models are 'creative.' They're statistical engines. Our job is to make them fast, reliable, and easy to steer toward your vision.
Every training clip comes from a stock library we pay for or a public-domain archive. If you create with Unionvora, you're not inheriting legal risk from scraped content.
Our support team are former editors and motion designers. They answer questions in under 15 minutes and actually understand temporal coherence and codec settings.
I started Unionvora because I was tired of watching great editors do boring work. AI should handle the grunt tasks—finding B-roll, resizing for platforms, generating draft animatics—so humans can focus on story and emotional beats. We're building tools I wish I'd had at the agency.