One-person product studio
A one-person product studio building independent products with AI, cloud infrastructure, and product systems.
I'm a programmer and AI Transformation engineer turning ideas into shipped products, starting with Clad.
Studio Snapshot
It is my independent product studio: a place where I design, build, launch, and grow products as a solo builder. I share the process publicly for builders who want to see what it actually takes to go from idea to product.
Idea to shipped product
I turn ideas into working products.
Signals to decisions
I learn from users, metrics, and product feedback.
Notes to distribution
I document decisions, failures, and lessons in public.
Current Product
AI personal color service
Clad helps people discover their colors, palettes, and beauty recommendations with AI. I built it because personal color analysis is useful, but often expensive, subjective, and hard to access.
Clad product mockup
Photo upload
Photo upload
A focused upload step that sets expectations before analysis.
Next action
Start with one photo and clear guidance.
Product proof
A focused upload step that sets expectations before analysis.
Product thinking
The goal is not to make an AI demo. The goal is to create a useful product experience: from photo upload, to analysis, to confidence, to recommendations people can actually use.
Personal color analysis can be expensive, subjective, and hard to access.
AI can make the first step more accessible and visual.
Raw AI output is not enough. Users need confidence, clarity, and next actions.
A good AI product helps the user decide what to do next.
Build Notes
I write about the process behind LumX: product decisions, AI workflows, cloud architecture, launch lessons, and what I learn from users.
LumX exists because I want to build products, operate them, and share the process with enough detail that other builders can learn from it.
Personal color analysis is useful, but often expensive, subjective, and hard to access. Clad explores how AI can make the first step more accessible and actionable.
Designing Clad means designing the product layer around AI: upload guidance, analysis flow, result explanation, palette visualization, recommendations, and trust.
AI Transformation is not just automation. It is workflow discovery, operating model design, evaluation, deployment, and continuous improvement.
Cloud, Kubernetes, and MLOps do not create a product by themselves, but production thinking helps a solo builder design reliable AI product systems.
AI Transformation Practice
It is operating model design.
My work sits at the intersection of software engineering, cloud infrastructure, AI systems, and product operations.
I think about AI Transformation as a systems problem: how teams discover workflows, design AI-native processes, evaluate outputs, deploy safely, and improve continuously.
Turning repetitive work into AI-assisted systems.
Designing agents, pipelines, evaluation, and feedback loops.
Building the infrastructure required to run AI products reliably.
Connecting user feedback, metrics, and iteration into a repeatable operating loop.
Engineering Proof
I combine product engineering, AI workflows, and cloud-native infrastructure to turn ideas into reliable products.
Case Study: Clad
Personal color analysis is useful but often expensive, subjective, and hard to access.
I designed an AI-assisted flow that turns a user photo into a color analysis and recommendation experience.
Frontend, image upload, model workflow, result generation, recommendation UX, analytics, and iteration.
Users want confidence, clarity, and next actions.
Operating Principles
Build small. Ship early. Learn from real users. Turn lessons into systems.
AI capability only matters when it becomes a useful product experience.
A solo builder needs repeatable systems, not one-off effort.
Real products, screenshots, metrics, and build logs matter more than big words.
Sharing the process creates accountability, distribution, and community.
Follow the build
I'm building LumX in public, sharing how I design, build, launch, and grow independent products from zero.