Most young students today face serious questions about their future. What should they study? What are their strengths? How can they build a meaningful career? Despite having access to more information than ever, finding answers to these life-defining questions remains a challenge.
The Learning Planet Institute (LPI) saw an opportunity to change that. Partnering with code.store, the nonprofit institute created Ikigai+, an AI-powered chatbot designed to guide students through four essential questions:
- What are you passionate about? (what you love)
- What natural talents do you possess? (what you're good at)
- What positive change can you create? (what the world needs)
- How can you sustain yourself? (what you can be paid for)
The result? Students now spend 45 minutes in deep conversation about their future, without being asked to.
LPI’s Quest to Inspire Gen Z
LPI explores new ways of learning that respond to the needs of youth and the planet. Through their workshops, they have already guided 2,500 students to discover their life purpose using Ikigai, a Japanese framework that explores what you're good at, what you love, what the world needs, and what you can be paid for.
The challenge LPI faced was scaling the framework to reach thousands more students without losing the personal experience. "The challenge was how to support the person with ideas and inspiration to help others if there is limitation with resources," explains Pascal Kolbe, Project Manager at LPI. With a €200,000 budget and a four-month timeline, they needed a solution that could scale fast and remain effective.
From Concept to Real Life
The journey of Ikigai+ began with a simple but crucial decision: rather than developing an entirely new platform, code.store would take the core idea of the project and bring it directly to where students already spent their time. "This project was created with the purpose of helping people find a sense of direction in their lives," shares Olga Zviezdicheva, who led the development team.
The initial weeks were spent observing how students interacted on Discord, the platform they were most familiar with. The team noticed patterns in communication, identified key features students engaged with, and pinpointed the times when they were most active. These insights were pivotal in shaping the technical approach.
Building the Foundation
The core solution required careful planning and innovative thinking. сode.store’s developers chose Python to build the system, as it was well-suited for handling complex conversations while maintaining user context. OpenAI’s natural language processing powered the bot, but instead of generating new responses each time, the developers implemented a vector-based storage system. This approach allowed the bot to quickly retrieve and adapt past successful interactions.
A significant innovation was the development of the custom Electro framework. "Before Electro, every small UX change needed a developer," explains Nikola, the lead developer. "Now content experts can modify the system directly." This advancement saved time and enabled rapid updates during the tight four-month timeline. Speed remained a top priority throughout development. "Students would lose interest if the AI took too long to respond," Nikola notes. To address this, the team optimized the bot by combining vector storage for common conversation paths with PoEditor for localization. Response times improved by 60% without compromising quality.
For image generation, the team stored vector versions instead of creating new ones repeatedly. "We wanted something sustainable," Pascal, the project manager, explains. "Both environmentally and in terms of system resources." The optimized solution significantly reduced server load while maintaining efficiency.
Testing and Iteration
The team ran two-week development sprints with student volunteers from Cergy University. This rapid iteration process proved crucial for transforming a complex personal development program into something students would actually use. Early feedback shaped every aspect of the experience. Students preferred shorter, more frequent interactions over long sessions. They wanted to save their progress and share insights with peers. Most importantly, they needed clear privacy controls they could trust.
Each insight led to improvement. "We originally planned longer guided sessions," Pascal notes, "but students showed us they wanted more control over the pace." The team rebuilt the conversation flow, making it more flexible and natural. They also added features for saving and sharing insights, always with privacy in mind.
Data security was foundational from the start. The team hosted all data in France under GDPR guidelines, creating an environment where students felt safe exploring personal topics. This trust translated into engagement: students return an average of three times per week, with many continuing their self-discovery journey for months.
The initial pilot with 500 Cergy University students revealed both successes and challenges. Students spent an average of 45 minutes per session exploring their purpose, but they also wanted more ways to share insights with peers."This course strengthened my skills and helped me find balance and purpose in life," shares Marie, who discovered her passion for sustainable design. Thomas, initially skeptical, adds, "The conversations opened my eyes to possibilities I hadn't considered." Sarah reflects, "Ikigai+ helped me set clear goals for my future."
Lessons from the Ikigai+ Journey
The Ikigai+ project offers practical insights for organizations planning AI-driven solutions:
Meet Users Where They Already Are
Building Ikigai+ on Discord eliminated the need to create a custom platform. Students were already familiar with Discord, which meant less friction and faster adoption. For teams, this is a reminder to leverage tools your audience already uses.
Make Security a Priority
From the start, Ikigai+ adhered to GDPR regulations, with all data hosted in France. This approach reassured students about their privacy. Planning for compliance and secure infrastructure upfront avoids issues later and builds trust with Gen Z users.
Optimize for Speed
Quick responses kept students engaged. Using vector-based storage for conversation paths and PoEditor for localization reduced lag by 60%. If your tool relies on real-time interaction, prioritize systems that can handle large-scale use without delays.
Enable Easy Updates
The Electro framework lets content teams make changes without developers, saving time during iterations. For AI projects, choose tools that allow non-technical users to tweak content and workflows independently.
Test, Listen, Improve
Starting with a 500-student pilot allowed the team to gather real feedback. Students wanted shorter interactions, progress tracking, and clear privacy controls. Continuous testing helped refine the experience to meet these needs.
What started with 500 students has grown to 1,400 participants, with plans to reach 40,000 more. The Learning Planet Institute is now developing MOOD (Massive Open Online Discovery), which will expand on Ikigai+'s success by combining AI-guided self-reflection with peer-driven learning communities. "MOOD represents everything we've learned about making complex ideas accessible through technology," Pascal shares. "It's about creating spaces where learning happens naturally, where AI enhances rather than replaces human connection."
The plan is to partner with universities like Cergy to scale quickly. Early feedback shows that 85% of students found the AI guidance helpful. The overwhelming response signals that MOOD can expand its reach and engage more participants. The success of Ikigai+ shows that meaningful AI engagement starts with understanding how your audience already uses technology. For organizations looking to drive meaningful engagement with AI, the path forward is clear: meet your audience where they are, build trust through transparency, and create systems that can evolve with user needs.
Want to discover how AI can transform user engagement? Subscribe to our newsletter and receive real-life case studies directly in your inbox!
