AI Solutions
In an era when people ask AI before the search bar, we build content AI will cite — and verify the results in numbers
People now ask AI tools like ChatGPT instead of searching directly on Naver or Google. AI presents the sources it consulted alongside its answers. Which content gets chosen as those sources is becoming the new standard of search competition.
Re:B Studios is a GEO (Generative Engine Optimization) technology company that designs content AI can reference and cite, and measures how often it is actually chosen.
The search engine era had SEO. The AI search era needs GEO.
Research grounding
This approach is not Re:B's claim alone. Research presented at ACM KDD 2024 reported that applying elements that increase credibility and extractability — sources, statistics, quotations — can raise content visibility inside generative engines by up to 40%.¹ Follow-up work found citation rates improved by an average of 17.3% when document structure was optimized,² and pages actually used in AI answers tend to be rich in directly citable evidence such as definitions, figures, and comparisons.³ Still, effects vary by engine and domain. That is why a system for measuring and verifying directly in each domain is needed.
Engine 1
AI content technology
① AI that creates content
In operationA production system that defines a virtual persona's personality, voice, and worldview as data, and controls AI to follow that identity consistently. Beyond text, we implemented a multimodal pipeline extending to avatars, motion photos, and short-form video output. Through our own publication Re:PAGE (ISSN 3140-2437), we verified the operability of production, review, and publishing.
② People who refine content
In operationWe never publish AI writing as-is. Every publication passes final approval by a three-person review board. We also regularly collect the official policy documents and public changelogs of search and AI platforms, and fold them — after board review — into new versions of our writing and review rules.
③ Data that accumulates and learns
UpgradingWe label produced content by persona traits, topic, and quality, building our own dataset. "Does this piece sound like this persona?" is also scored by an AI evaluation system, while final publication is always decided by the three-person review board.
④ AI that measures citation
Target: Q4 2026Initially, about 50 domain queries run periodically across two generative search engines, tracking which content gets cited as sources. We deliver as our own report a number that is not yet standardized and has been hard to check consistently by domain.
Engine 2
App & service development capability
We implement technology as real, operating services.
A company that turns AI technology into working services instead of leaving it as ideas.
What this team has actually done
Built & operated an AI search engine
Launched an app with search and AI summarization as a solo developer — over 10,000 downloads without ads and 6,500 monthly users.
(member's personal project)
Operated large-scale data pipelines
Built and ran a pipeline auto-collecting data on 100,000+ influencers, processing 500,000 items per day.
(member's work at a previous employer)
Ran payment services reliably
Built a payments-enabled service from scratch and operated it at 99.9% uptime, 150,000 transactions per month, and a 0.02% error rate.
(member's work at a previous employer)
These experiences are now combined into a single team capability inside Re:B Studios.
Roadmap
| H1 2026 ✓ | AI content engine in production · Re:PAGE first issue · in-house app development completed · in-house AI diagnostic system built & operating |
|---|---|
| Q3 2026 | Official app launch · AI dataset systematization |
| Q4 2026 | GEO citation tracking system v1 live |
| H1 2027 | AI-search-ready content serving system · specialized AI model research begins · content-linked micro-tool services launch |
| H2 2027 | GEO Scorer launch — predicting a piece's AI citation potential before publishing · AI persona chat development · 'Living Magazine' development |
- ¹Aggarwal et al., "GEO: Generative Engine Optimization," Proc. ACM SIGKDD (KDD '24), 2024, pp. 5–16.
- ²Yu et al., "Structural Feature Engineering for Generative Engine Optimization," arXiv:2603.29979, 2026.
- ³Zhang et al., "From Citation Selection to Citation Absorption," arXiv:2604.25707, 2026.
For companies that need systems — or wonder if AI cites them
From system planning, development, and operations contracts to joining the GEO citation report pilot — we welcome your inquiry.