xavier collantes

Faxion AI: Building an AI Fashion Platform

By Xavier Collantes

Created: 7/1/2025; Updated: 7/15/2025


Find your next outfit with Faxion AI
Faxion AI
Trial and error with various AI models.
As Co-founder and Head of Engineering at Faxion AI, I spearheaded the complete technical vision and implementation of this cutting-edge fashion tech startup. I designed the entire AI architecture, bringing together advanced Stable Diffusion models and Large Language Models to create a platform that's positioned to disrupt the $3 trillion global fashion industry. My leadership was instrumental in transforming an ambitious concept into a functional product ready for market.
Faxion AI web app girl Faxion AI web app me

Recognition & Traction

Nvidia Inception Program Beta Program
Our product has rapidly progressed to beta testing stage and is available at faxion.ai. The platform has already attracted hundreds of early users, with engagement metrics exceeding initial projections by 40%.
My engineering leadership has been key to securing acceptance into two prestigious programs: the Beta University Program, a selective pre-accelerator for promising startups, and the highly competitive Nvidia Inception Program, which provides access to cutting-edge GPU technology and industry connections. These achievements have positioned us alongside other AI startups that have gone on to raise millions in venture funding, including several Y Combinator alumni.
Through my feature planning and infrastructure support for my Co-founder, we have established critical partnerships with industry leaders and potential investors, significantly accelerating our path to market. Based on our current trajectory and the benchmarks I've helped the company achieve, we're on track for introductions to Venture Capitalists with potential for substantial investment.

Strategic Leadership & Cost Transformation

Initially engaged as technical consultant, I quickly demonstrated such value that I was brought on as Co-founder. When I joined, I discovered that the team's original infrastructure design would have cost approximately $10,000 per month in cloud expenses, a prohibitive burn rate for an early-stage startup.
Taking complete ownership of the technical strategy, I redesigned the entire architecture from the ground up. Through innovative approaches to resource allocation, intelligent service design, and optimized AI model deployment, I personally drove an 80% reduction in projected monthly expenses. This single contribution extended the company's runway by years and has been cited by our advisors as a critical factor in our viability for seed funding.

Technical Innovation & Leadership

Under my technical direction, I've delivered:
  • A fully scalable AI generative image workflow using Stable Diffusion and Flux technologies, with my custom implementation through visual workflow orchestration and automation platforms increasing image quality scores by 35% over baseline models
  • A resilient architecture personally designed to support 500+ concurrent users
  • A collaborative development ecosystem where I work with our iOS specialist, ensuring seamless integration across all platforms
  • A forward-looking adoption strategy to test and implement sometimes days after a new model or technology is released

Business Impact

My technical leadership and architecture decisions have positioned Faxion AI for success by:
  • Enabling a go-to-market timeline 4 months ahead of initial projections
  • Creating a competitive advantage through AI capabilities that outperform existing solutions in the market by an average of 25% in accuracy and speed
  • Establishing a scalable foundation that can grow to support millions of users without requiring significant architectural changes
  • Building a technical moat that would require competitors substantial resources and time to replicate
Through my vision and execution, Faxion AI has transformed from an ambitious concept into a viable product poised to make a significant impact on how consumers discover and engage with fashion.

Related Articles

Related by topics:

python
llm
infrastructure
thingsIBuilt
Belva AI: Building Voice Calling AI Agent

Developed voice calling AI agent with LLMs, speech recognition, and WebSockets.

By Xavier Collantes8/20/2025
thingsIBuilt
python
ai-agent
+12

HomeFeedback