SEM
SEM is an AI-powered platform designed for video editors, content creators, and creative teams who need to find, generate, edit, organize, and reuse sound effects more efficiently. The platform aims to simplify the audio workflow behind video creation by allowing users to generate sound effects from prompts, create audio for video content, manage their own sound library, and potentially share or reuse generated assets inside a clean creative workspace.
Visit Landing Page
My Contributions
- I am building SEM on my own, handling the full-stack development, AI integrations, and product design.
- I am also responsible for marketing and everything related to the startup, including strategy, growth, and overall direction.
Tech Stack
- Next.js & TypeScript - Frontend application, user interface, routing, and interactive product experience
- FastAPI - Backend APIs, business logic, authentication, and generation workflows
- PostgreSQL - Persistent storage for users, audio assets, jobs, credits, and metadata
- Redis & Celery - Asynchronous job queues for long-running audio and video processing tasks
- AI Audio Providers - Sound generation, audio transformation, and provider-based model integration
- ffmpeg - Audio and video processing, mixing, exporting, and media conversion
- S3 / MinIO - Persistent object storage for generated audio, uploaded files, previews, and final exports
- Stripe - Credit purchases, subscriptions, and billing workflows
- Docker - Containerization and deployment preparation
- Pytest & TypeScript checks - Backend and frontend testing, validation, and build reliability
Features
- Text-to-sound-effect generation
- Video-to-audio workflow
- AI-assisted sound effect creation for video projects
- Prompt-based audio generation
- User sound libraries
- Audio preview and download system
- Public sound exploration interface
- Credit-based generation system
- Audio publishing
- Asynchronous generation jobs
Project Status
Ongoing - SEM is currently in active development as my first startup aimed at becoming a real product. I am focusing on validating the core idea, improving the user workflow, and understanding the legal and licensing aspects of AI-generated audio before launch. Despite being early-stage, it is already a complex full-stack project combining AI, audio processing, video workflows, and backend infrastructure.
Project Screens
Short placeholders for now. Replace these notes with a fuller explanation later.
Explore Sounds
Main discovery view for browsing community-created sound effects, searching by need and opening sound cards with metadata such as duration, category and engagement.
