Use Case Intelligence RSS Search use cases Search
OpenAI Software / SaaS Asia-Pacific & Oceania

Wrtn

Marketing Personalization

Wrtn needed to move beyond narrow productivity tools and make AI feel natural, localized, and useful for everyday Korean users across creativity, learning, and conversation at mass scale.

Read the original story Explore all use cases
10% jump in month-one retention
15% increase in session time after a router upgrade
6.5 million monthly active users
8% increase in daily active users within a week of GPT-5 launch

Reported impact

Business outcome

Wrtn reached 6.5 million monthly active users, doubled retention, quadrupled time spent per user, and used same-day GPT-5 adoption plus a modular model router to keep improving consumer engagement and localization quality.

  • Customer Experience
  • Revenue Growth
  • Scalability

Solution

Wrtn built a consumer AI stack that combines localized prompting, memory scaffolding, and model routing so millions of users can interact with AI in natural Korean rather than translated English patterns. Its Crack character-chat product and broader suite of AI Supporter and Labs experiences sit on top of a router architecture that sends lightweight classification and traffic-direction tasks to GPT-4o mini and GPT-4.1 mini, while heavier tutoring, counseling, and multimodal interactions run on GPT-4.1, GPT-5, and audio models. Persona prompts, role guidance, and temperature tuning shape the tone and pacing of responses, while each new OpenAI release can be swapped into the stack with minimal disruption. That architecture let Wrtn improve retention, time spent, and localization quality while scaling lifestyle AI across Korea and into new East Asian markets.

Data flow

User requests enter Crack or other Wrtn apps, where the router architecture classifies intent with smaller models and sends heavier work to GPT-4.1, GPT-5, or multimodal audio paths. Persona prompts, memory scaffolding, and localization layers shape the response before the product returns text or voice output to the user.

Solution architecture

7 components · 3 layers
  1. Compute
    • GPT-4o mini Handles lightweight routing and classification tasks in the product stack.
    • GPT-4.1 mini Supports low-cost classification and traffic-direction tasks in the modular architecture.
    • GPT-4.1 Runs heavier tutoring and reasoning tasks where better context handling and localization matter.
    • GPT-5 Improves consistency, instruction adherence, personalization, and overall consumer engagement.
  2. Serving
    • Crack Serves as Wrtn's open-ended character chat product for storytelling, language learning, and self-expression.
    • multimodal TTS Adds spoken tutoring and other audio interactions to Wrtn's consumer AI experiences.
  3. Orchestration
    • router architecture Classifies requests and directs them to the most appropriate model path for cost, latency, and capability.

Architecture clues

  • Heavy tasks are separated from lightweight classification for cost control
  • Localization tuned for Korean slang and cadence
  • Persona-based prompts and memory scaffolding shape responses
  • Router architecture swaps in new models with minimal disruption

Evidence from the source

"After a router upgrade, session time rose 15% and month-one retention jumped 10%."
"That work led to Crack, Wrtn's open-ended character chat platform"
"Today, Wrtn's product suite spans AI Supporter, Labs, and Crack, reaching 6.5 million monthly active users."
"Under the hood, Wrtn runs a router architecture"
Transparency Evidence & methodology 100/100 reference value · 87% confidence · 1 source

Review how this reference was scored and trace each normalized claim back to its documented source.

01

Scoring methodology

How scores are calculated

Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.

Customer reference value100/100
  • Problem clarity20/20
    Measures whether the customer's operational problem is specific enough to compare.
  • Implementation depth25/25
    Rewards named components, data flow, integrations, and architecture detail.
  • Outcome specificity20/20
    Measures whether the reported benefit is concrete rather than a generic claim.
  • Quantified impact25/25
    5 metrics describe measurable impact.
  • Deployment scale10/10
    Rewards evidence of a live or scaled deployment and meaningful integration scope.
Evidence confidence87%
  • Source authority20/25
    The main evidence is an official vendor-published customer story.
  • Claim traceability21/25
    10 of 12 key fields link to a named source.
  • Direct evidence20/20
    4 direct source quotes retained.
  • Date provenance8/10
    The publication date comes from publisher-controlled metadata.
  • Source transparency8/10
    1 source reference document publisher, URL, authority, retrieval, and supported fields.
  • Implementation specificity10/10
    Rewards source-backed components, data flow, quantified impact, and a specific outcome.
02

Traceability

Sources & provenance

Field coverage shows which facts each source supports. Dates are shown only when documented by the source.

  1. OpenAI

    Wrtn customer story

    First-party
    https://openai.com/index/wrtn/
    Source type
    Vendor-published customer story
    Published
    Oct 2, 2025
    Supports
    • Customer
    • Business problem
    • Solution
    • Products
    • Solution components
    • Data flow
    • Integrations
    • technical area
    • use case category
    • Business outcome
    • outcome category
    • Metrics
    • Architecture
    • evidence quotes
    • Publication date

Date evidence

Not separately sourced Industry, Region, Company size. Treat these normalized attributes as lower-confidence until a dedicated source is added.

Related

Similar use cases

See all like this
OpenAI Aug 10, 2026 Zapier Software / SaaS North America At Zapier, the enterprise marketing team is using ChatGPT Work to build automated funnel optimization systems, enabling them to deliver more value to their sales team with less manual effort. ChatGPT Work operates… Now, the enterprise marketing team can perform automatic QA/QC on thousands of leads every month. The… IT Operations AutomationMarketing Personalization ChatGPT Work OpenAI Aug 3, 2026 Circles Software / SaaS Asia-Pacific & Oceania Circles needed to turn rich telco signals such as usage, billing, network activity, and location into real-time support and monetization actions instead of reactive, siloed customer journeys. 22% increase in ARPU in Singapore from AI-driven personalization Customer Service AutomationMarketing Personalization AI Concierge · CareX · Codex OpenAI Jan 26, 2026 Indeed Software / SaaS Global Indeed needed to expand AI across job search, employer hiring workflows, and internal operations while keeping human decision-makers in the loop and building employee confidence in how to use the technology well. About 70% of sponsored applications come from AI-powered recommendations Customer Service AutomationMarketing Personalization Career Scout · OpenAI APIs · Smart Sourcing