AI News
MIXI ChatGPT Enterprise Deployment Achieves 90% Work Hour Reductions
MIXI deployed ChatGPT Enterprise across 1,000+ employees in 45 days, creating 1,600+ custom GPTs with some projects achieving over 90% work hour reductions.
MIXI operates popular mobile game "MONSTER STRIKE" and family photo-sharing app "FamilyAlbum" with 25 million users across 175 countries. The company initially subsidized ChatGPT Plus subscriptions for over 1,000 employees before transitioning to the enterprise platform for unified data handling and security controls.
Rapid 45-Day Enterprise Rollout Strategy
MIXI's deployment timeline contrasts with typical enterprise software implementations that span multiple quarters. The company partnered directly with OpenAI's Customer Success team to address departmental concerns around personal data handling and operational continuity through established mitigation playbooks.
Three core enablement programs supported the rollout:
- ChatGPT 101 training sessions for all employees focusing on practical daily prompts
- New-hire workshops for April 2025 graduates building AI fluency from onboarding
- Hackathon events where engineers built applications using OpenAI's Agents SDK
Tatsuma Murase, Director and Senior Corporate Officer, noted the security advantages: "We needed to check each employee's AI settings individually, and our rules for using company data were unclear. It was obvious we had to create a secure, unified environment."
Custom GPT Development for Brand Consistency
MIXI's FamilyAlbum division developed specialized GPTs for advertising and brand management, reducing creative planning work by approximately 28 hours monthly. The team built two primary tools:
The Copy Checker GPT reviews marketing taglines to ensure brand voice consistency, suggesting softer language when proposals include aggressive wording. The Creative Planning GPT enables non-native speakers to generate localized ad copy respecting both brand values and regional nuance.
Nobuhisa Hirata, Assistant Chief General Manager of FamilyAlbum Business, explained the collaborative approach: "Our teams have developed GPTs that encourage a collaborative culture, where members actively test and refine prompts together."
Investment Analysis Automation Results
MIXI's investment division, which monitors over 3,000 portfolio companies across 70+ funds, developed a "VC Fund Initial Review Support" GPT for proposal evaluation. The tool reduced individual investment review time from 1-2 hours to 5-10 minutes while maintaining evaluation quality consistency.
Gohshi Araki, Head of Corporate Development, emphasized the strategic shift: "The capabilities provided by AI were indispensable. Now, we use ChatGPT for investment evaluations and monitoring, significantly aiding risk assessment." The division reports workload reductions exceeding 70% in multiple use cases, with saved time redirected toward relationship-building activities.
Enterprise Adoption Metrics and Implications
Within three months, over 80% of MIXI employees became active weekly ChatGPT users, creating more than 1,600 custom GPTs. This adoption rate suggests strong organizational AI readiness when supported by structured deployment and training programs.
For European enterprises evaluating similar deployments, MIXI's approach highlights the importance of dedicated customer success partnerships and comprehensive training programs. The 45-day timeline demonstrates feasibility for rapid organization-wide AI tool adoption when security and data handling concerns receive upfront attention.
CTO Junpei Yoshino summarized the productivity philosophy: "When operational efficiency rises, we win back time for imagination. And with that time we use AI again, accelerating the next idea."
Original source: OpenAI MIXI case study published by OpenAI.
AI News Updates
Subscribe to our AI news digest
Weekly summaries of the latest AI news. Unsubscribe anytime.
More News
Other recent articles you might enjoy.
Chat with 100+ AI Models in one App.
Use Claude, ChatGPT, Gemini alongside with EU-Hosted Models like Deepseek, GLM-5, Kimi K2.5 and many more.