Navora: AI-Driven Learning
A research-led mobile application designed to optimize reading efficiency and information retention. By merging cognitive psychology principles with Generative AI, Navora transforms how users interact with complex information, serving as a personal case study for high-impact educational technology.
/ Project Milestones
The journey and phases to bring this project to life.
Product Founder & Architect
Personal Research Project
I initiated this project to bridge the gap between cognitive psychology and digital learning[cite: 40]. The strategic core was to design an experience that respects mental load and focuses on deep processing[cite: 40]. I mapped out user flows based on linguistic structures and information retention theories, ensuring that every technical decision—from the choice of Flutter for a seamless UI to the backend architecture—served a specific pedagogical purpose. This phase was about defining a product that doesn't just display text, but actively facilitates learning through intentional design.
Lead Developer
Personal Research Project
I strategically integrated Generative AI (LLMs) not as a gimmick, but as a core utility for automated content prototyping and personalization[cite: 41, 38]. By implementing prompt engineering and serverless functions, I created a system that can summarize, simplify, and generate retention-focused exercises in real-time. This decision allowed for a 'lean' content strategy where the AI acts as an invisible tutor, adapting the material to the user's reading level[cite: 41]. The focus was on optimizing development velocity using AI tools to build unit tests and documentation, allowing me to maintain high code quality as a solo developer.
Full-stack Developer
Personal Research Project
To increase user engagement without compromising the serious goal of reading efficiency, I implemented a gamified ecosystem using Rive for high-performance animations[cite: 42]. I made the strategic choice to use an 'ads-for-lives' monetization model restricted only to game levels, keeping the core 'Practice' utility direct and friction-free. This balanced approach ensures that the educational value remains the priority. The architecture uses Firebase for real-time lead collection and user progress tracking, creating a data-informed feedback loop that allows for continuous refinement of the learning algorithm based on actual user behavior.
Product Founder & Architect
Personal Research Project
I initiated this project to bridge the gap between cognitive psychology and digital learning[cite: 40]. The strategic core was to design an experience that respects mental load and focuses on deep processing[cite: 40]. I mapped out user flows based on linguistic structures and information retention theories, ensuring that every technical decision—from the choice of Flutter for a seamless UI to the backend architecture—served a specific pedagogical purpose. This phase was about defining a product that doesn't just display text, but actively facilitates learning through intentional design.
Lead Developer
Personal Research Project
I strategically integrated Generative AI (LLMs) not as a gimmick, but as a core utility for automated content prototyping and personalization[cite: 41, 38]. By implementing prompt engineering and serverless functions, I created a system that can summarize, simplify, and generate retention-focused exercises in real-time. This decision allowed for a 'lean' content strategy where the AI acts as an invisible tutor, adapting the material to the user's reading level[cite: 41]. The focus was on optimizing development velocity using AI tools to build unit tests and documentation, allowing me to maintain high code quality as a solo developer.
Full-stack Developer
Personal Research Project
To increase user engagement without compromising the serious goal of reading efficiency, I implemented a gamified ecosystem using Rive for high-performance animations[cite: 42]. I made the strategic choice to use an 'ads-for-lives' monetization model restricted only to game levels, keeping the core 'Practice' utility direct and friction-free. This balanced approach ensures that the educational value remains the priority. The architecture uses Firebase for real-time lead collection and user progress tracking, creating a data-informed feedback loop that allows for continuous refinement of the learning algorithm based on actual user behavior.