Marwand Ayubi

Founder & Architect building the future of Healthcare & Enterprise Tech

End-to-end product delivery. I wrote the first line of code for Vocadoc and shipped it to production. From architecture to deployment, from ML pipelines to frontend polish.

Real-world infrastructure at scale. I manage the Neolyth datacenter, run self-hosted Proxmox + Kubernetes clusters, handle AWS deployments, and orchestrate zero-downtime releases with full observability.

AI/ML in production. PyTorch + TensorRT speech-to-text, prompt engineering, GPU-accelerated inference, continuous model evaluation. Not just theory - real systems processing real medical consultations.

Security & compliance. Encrypted data flows, RBAC, audit logging, GDPR + ISO 27001 alignment, HIPAA-ready design. Healthcare data demands it, and I deliver it.

If a problem falls outside my stack, I learn it fast and ship.

stack = {
    "backend": ["Python", "FastAPI", "Flask", "Node.js"],
    "frontend": ["React.js", "Next.js", "JavaScript"],
    "ml_ai": ["PyTorch", "TensorRT", "Prompt Engineering", "GPU Inference"],
    "infrastructure": ["Docker", "Kubernetes", "Proxmox", "Self-Hosted Datacenters"],
    "devops": ["GitLab CI/CD", "Zero-Downtime Deploys", "IaC", "Observability"],
    "databases": ["PostgreSQL", "Redis", "TimescaleDB", "Vector DBs"],
    "cloud": ["AWS", "Hybrid Cloud", "Multi-Cloud"],
    "security": ["Encryption", "RBAC", "Audit Logging", "GDPR/ISO 27001"]
}

A decade of hands-on work. From first commit to live rollout.

Note: This stack represents my primary tools, but my expertise extends far beyond these technologies. From embedded systems to cloud platforms, from frontend frameworks to AI inference engines. If it's in the technology domain, I've likely worked with it.

Important Context: This isn't my primary development platform. At Neolyth, we run self-hosted GitLab for superior code isolation and security. Most of my daily work lives there.

What you'll find here:

• Open-source utilities and tools I've built
• Public scripts and automation helpers
• Experimental projects and proof-of-concepts
• Community contributions

My commit graph doesn't reflect my actual development velocity! Judge the code quality, not the activity metrics.

When my schedule allows, I advise teams on applying AI, automating workflows, and architecting scalable cloud or self-hosted solutions. I own Vocadoc's roadmap, certifications, and growth while steering Neolyth's wider push into AI software.

If you're building something in healthcare tech, AI infrastructure, or need someone who can navigate from hardware to application layer - let's talk.

I build AI products and use AI tools in my workflow. Both require the same principle: human-centric design with critical evaluation.

AI should assist and augment, not replace judgment. Whether I'm architecting Vocadoc's clinical documentation system or using AI to accelerate my own development, I approach it with the same rigor - understanding the output, validating the logic, and maintaining accountability.

The technology is transformative when applied thoughtfully. The key is knowing when to leverage it and when to step back and understand the fundamentals yourself.