Labhakar

We were driven by shrinking attention spans and the rise of superficial online content, seeing a need for quick, engaging ways to learn complex topics. We aimed to combat "brainrot" and information overload by providing instant, valuable learning bites. This led to the idea of generating educational videos on demand.

LabHackar instantly transforms any typed topic into a complete multimedia lesson using AI. Our platform generates a video with synchronized narration, subtitles, and relevant visuals in moments. It makes learning complex subjects accessible and engaging through personalized, compact video content.

We connected several APIs, starting with Gemini integrated into our website for user input. Vertex AI and Gemini generate the script and visual concepts, which are then passed to Shotstack. Shotstack handles the cloud-based rendering, assembling the AI narration, visuals, and subtitles into the final video.

Our main hurdle was integrating and adapting the different APIs (Vertex AI, Gemini, Shotstack) to work together seamlessly. Passing data correctly between them was complex, often requiring significant trial-and-error due to limited specific code examples. Making distinct services communicate effectively proved challenging.

We successfully built a functional end-to-end pipeline that takes a simple text prompt and delivers a full multimedia video lesson. Overcoming the integration challenges and orchestrating Vertex AI, Gemini, and Shotstack to work together reliably is our key accomplishment. We are proud to have brought this complex concept to life.

This project provided deep, practical experience in multi-stage API integration, working directly with advanced AI and video rendering services. We significantly improved our skills in system design for distributed services and cross-platform debugging. Ultimately, we learned the value of persistence when tackling novel technical integrations.

We plan to continue actively developing LabHackar, focusing on refining the AI content quality and user experience. Our long-term vision is clear: we aim to develop this project into a full-fledged startup. We believe AI-driven micro-learning videos have the potential to revolutionize education access.

Built With

• api
• gemini
• github
• node.js
• shotstack
• typescript
• vertex