Lena Hoffmann is a Berlin-based independent AI developer with a research-engineering background at Hugging Face. After two years inside the multimodal model trenches she left academia-flavored work to build small, focused products that let everyday users actually feel what Gemini's image, video, and audio APIs can do. GeminiOmni started as a weekend stack of demos that went viral on Twitter — now it's her full-time work, shipping one Gemini-powered tool at a time.
What I work on
GeminiOmni is my independent project: a multimodal toolkit wrapping Google Gemini, Veo 3.1, and Nano Banana into tools normal users can actually use — text-to-video, image editing, 1M-token PDF chat. One workspace, one credit balance, pay-per-generation. No subscription.
Background
- Multimodal AI applications
- Google Gemini API integration
- AI product engineering
- Generative AI UX design
- Indie AI tool development
Contact
Email: [email protected]