// saas, shipped

NIUJU

An AI portrait studio, designed, built and shipped solo with AI tooling. Security and privacy were part of the plan from the first commit.

[ SAAS ][ SHIPPED ][ SECURITY BY DEFAULT ]

Try NIUJU live

niu-ju.com

NIUJU homepage: the headline 'AI. Photography. Studio.' next to sample portraits of synthetic people, each labelled AI-generated

About the product

NIUJU (niu-ju.com) is an AI photo studio built and run by XIVU, a software studio in Fürth, Germany. Users upload up to four photos of themselves, pick a style and get AI-generated, studio-style photos. Customers buy one-time credit packs from €29 incl. VAT, or gift cards, through Stripe. There are no subscriptions.

The problem

People want professional-looking photos for LinkedIn, CVs and social profiles without booking a photographer or a studio. To get them, they hand over photos of their own face. That makes trust part of the product: the data handling has to be clear, short and true.

The brief

Ship a real, billable product, not a prototype. Solo, with AI tooling standing in for a full engineering team, and with privacy and security built in from the first commit.

Credits and scope of work

Year: 2025-2026

Built solo by XIVU

Product strategy
Design system
Frontend
Supabase architecture
Stripe payments
Privacy & security engineering
AI pipeline

What was built

[ STUDIO ]

A guided studio flow that turns a few inputs into consistent, on-brand portraits.

[ CREDITS ]

A credit system that meters generation fairly and ties cleanly to billing.

[ PAYMENTS ]

One-time credit top-ups and gift cards through Stripe. No subscriptions, and we never store card data.

[ DELETION ]

A deletion system that honours erasure requests across the whole data model, not just one table.

The hard parts

[ PERSONAL PHOTOS ]

People upload photos of themselves. They are never used to train AI and are deleted automatically 7 days after upload.

[ DATA LOCATION ]

Photos and results are stored in the EU (Ireland). The AI step can run outside the EU, mainly in the US, under a data processing agreement with EU Standard Contractual Clauses.

[ PROMPT SECURITY ]

Guardrails that make it hard to steer the generation pipeline into misuse or data leakage.

// security and privacy

Security and privacy decisions

Data location. Accounts, photos and results are stored on EU servers in Ireland. The AI step can run outside the EU, mainly in the US, under a data processing agreement with EU Standard Contractual Clauses. We say so up front.
Encryption and access. Data is encrypted in transit (TLS) and at rest, and access to it is limited by access controls.
Deletion by design. Uploaded photos are deleted automatically 7 days after upload. Generated images are kept for up to 7 days for download, then deleted. Users can delete both earlier at any time, deleting the account removes them immediately, and daily encrypted backups are purged within 7 more days.
No training on your photos. Photos are never used to train AI. Our AI provider is bound by a data processing agreement and does not use them to train its models.
Payments. Payments run through Stripe, and we never store card data. Credits are one-time, with no subscriptions.
Reviews before release. Every change that touches personal data, accounts or payments passes security, privacy and payment reviews before it goes live.
Honest claims. NIUJU holds no ISO 27001 or GDPR certification, and we do not suggest otherwise. It creates new photos from the ones you upload; it does not recognise, identify or match faces.

// what it shows

What NIUJU shows about how we build

NIUJU runs on the same secure SDLC we use for client work. Data flows and access rights are planned first. The build collects only the data a feature needs and keeps it only as long as it is needed. Changes that touch personal data, accounts or payments are reviewed before release. The rules and decisions are written down, so the AI coding agents that do the build follow them in every session.

It was built solo by Chan Vu Luu with AI tooling. The agents write the code; product decisions, security and privacy trade-offs and the final release stay with him.

Secure Launch: a product built this way

// stack