UPDATE | The white-label version of the app is available! from sticky notes to production: how we wibecoded a multi-agent AI app for an eco-hair salon

New Architecture Live: Universal Blender Engine Now Available

we have taken the next step in fusing AI with blockchain. i have developed the Universal Blender Engine—a mathematical and logical synthesis engine designed to balance resources, ratios, and variables across any domain.

this is a white-label architecture. while it is currently pre-configured to showcase its capabilities through cosmetic formulation (calculating safe blending ratios for C-retinol and base creams), its true power lies in its total adaptability. you can inject your own industrial materials, team time-management parameters, or local energy community resources into its background logic, and let the engine optimize them.

how to access?

access to the core engine is secured by web3 infrastructure:

• License: A one-time, wallet-bound lifetime access key.

• Price: 20 USDC (Note: you will need a tiny fraction of WLD or ETH network token in your wallet to cover standard blockchain gas fees).

connecting your wallet and minting the license instantly unlocks the expert assistant and the core synthesis matrix. welcome to experiencing how the future of resource management is built.

for more information, read the storyline .

for the purchase of the license click and follow the guidance.

introduction of the case study

what happens when you pair a visionary eco-hair salon entrepreneur with the cutting-edge paradigm of AI-assisted development? you get Nieida—a sophisticated, production-ready digital mentor and formulation app that proves small business digitalization is the ultimate playground for advanced AI systems.

this is the story of how we bypassed traditional engineering bottlenecks to build a multi-language, multi-agent platform from scratch using the method we call wibecoding.

the co-design workshop: mapping analog wisdom to AI guardrails

every great application starts with empathy, not code. our journey began with a focused collaborative workshop with the entrepreneur to translate decades of holistic hairdressing expertise, Finnish folk traditions, and strict ESG (Environmental, Social, and Governance) principles into structured machine logic.

over a series of iterative alignment sessions, we mapped out the core identity of the Nieida application:

• the brand archetype: a precise blend of 50% The caregiver (nurturing, supportive), 30% The everyman (grounded, relatable), and 20% the creator (crafting unique, organic formulations).

• the operational tone: professional, fluff-free, and encouraging, operating natively across finnish, swedish, and english.

• the code of conduct: deep respect for nature, integration of local botanical knowledge (like nettle and birch leaf rinses), and explicit customer affirmation.
  
  

the architecture: under the hood of the eco-salon mentor

Nieida isn't just a simple wrapper around a chatbot; it is a highly specialized multi-agent application that runs automated physical analysis, mathematical color forecasting, and strict safety filtering.

1. computed vision & hair porosity analytics

to formulate the perfect botanical dye, Nieida requires a snapshot of the client's hair via an integrated device camera. the application analyzes image contrast and texture.

high contrast and surface frizziness mathematically translate to high porosity, automatically alerting the orchestration layer to shorten the processing time because porous hair absorbs plant pigments rapidly.

2. the Kubelka-Munk formulator

at the heart of the calculation layer sits a mathematical model traditionally used in paint and textile formulation. nieida applies a variation of the Kubelka-Munk theory to calculate the ideal ratio of Henna (H), Indigo (I), and Cassia (C) required to reach a target shade based on the client's analyzed starting base color. [the version 2.0 of app has enriched with new ingrediends].

the relationship between surface reflectance and pigment concentration is governed by the following scientific framework:

• K is the absorption coefficient of the botanical blend.

• S is the scattering coefficient of the hair structure.

• R is the light reflectance measured from the hair sample.
  
  

3. the toxin filter & the safety agent

in the eco-salon world, client safety is non-negotiable. Nieida features an uncompromisably strict Toxin Filter that intercepts any system suggestions containing synthetic or toxic ingredients (such as paraphenylenediamine, or PPD).

sitting right beside the calculator is a dedicated safety agent running automated validation logic:

expected output	trigger condition	system action
PASS	healthy, untreated hair seeking a warm copper tone.	generates standard Henna/Cassia recipe; logs zero risk flags.
WARNING	bleached or highly porous hair attempting a cool dark brown/black.	flags high risk of a greenish tint or excessive darkening; demands pre-pigmentation and a strand test.
FAIL / CRITICAL	cross-reactive botanical combinations or unexpected structural threats.	halts formulation instantly, triggers an immediate first-aid fix, and escalates to physical packaging verification.

the wibecoding stack: iterating at the speed of thought

building this complex architecture did not require an enterprise development team. instead, the entire system was brought to production-ready status via wibecoding—leveraging high-level conversational AI architectures to scaffold, debug, and ship features continuously.

the lean, ultra-efficient tech stack consisted of four pillars:

[ Google AI Studio ] ──> [ Gemini 2.5 Flash ] ──> [ GitHub ] ──> [ Netlify Production ]
      

• Google AI Studio: the playground for prompt engineering. for stress-testing the system instructions, ensuring the safety agent caught high-indigo risks on porous hair long before writing frontend components.

• Gemini 2.5 Flash: chosen for its lightning-fast processing, deep context window, and exceptional multimodal capabilities. it seamlessly handles text chat, structures tabular responses, and processes camera images simultaneously.

• GitHub: collaborative safe haven. by separating the volatile, frequently changing conversational product catalog (knowledgeBase.ts) from the mission-critical formulation algorithms (formulatorKnowledge.ts), we could push updates safely without breaking the core mathematical calculations.

• Netlify: continuous deployment at its finest. every change pushed to the main repository was compiled, audited by the TypeScript compiler, and served live globally in under twenty seconds.

why this matters for the future of knowledge work

the creation of Nieida highlights a massive shift in how small business digitalization happens. we didn't just build a database of products; we captured the living, breathing expertise of an artisan industry and reinforced it with hard computational math and safety guardrails.

for knowledge workers and tech enthusiasts, the takeaway is clear: the barrier between a domain expert’s intent and a live software solution has completely vanished. when domain expertise guides the vision, and AI orchestration handles the heavy lifting, enterprise-grade tools can be manifested straight out of a local boutique environment.