shipped
iOS and Android app with headphone integration
An iOS and Android app with speech-to-text, an LLM chat, and translation. We also connected it to the headphone manufacturer's firmware.
> 2 months to build
> shipped to the App Store and Google Play ourselves
> handled business verification
iOSAndroidSpeech-to-TextLLM ChatTranslationFirmware Integration
shipped
Telegram AI bot for outbound sales
The bot reaches out to prospects in Telegram, handles the conversation, and moves them toward a sale. Anti-detect setup, account rotation, runs autonomously.
> 100+ leads per day
> no human in the loop
> 4 weeks to build
PythonNext.jsPostgresRedisTelegram APIAnti-detect
shipped
SEO article factory for a top-5 online university
An automated pipeline for psychology SEO content. It covers topic selection, drafting, fact checking, and publishing straight to WordPress.
> 200+ articles delivered
> about 800 hours saved
> built in 2 weeks
Pythonn8nWordPress APIOpenAIRAG Pipeline
shipped
SEO content for a network of conscription websites
The system generates articles for a network of websites about military conscription. Before publishing, it verifies claims against Russian legislation, the official Schedule of Diseases, and other primary documents instead of recycled web copy.
> built in 2 weeks
> fact-checked against primary sources
> content for multiple websites
SEOFact-checkRussian LegislationSchedule of DiseasesOfficial Sources
shipped
AI nutritionist in Telegram
Send a food photo, the bot counts calories, builds a meal plan, and keeps a diary. Built-in subscription model.
PythonPostgresRedisTelegram
shipped
Fanfy Studio — AI fanfic generator
Users choose characters, setting, and plot beats, then AI writes the full story. Web app with streamed generation.
> 4 weeks from zero to prod
Next.jsTypeScriptFastAPIPostgresOpenAIStreaming
shipped
MechaPost — all-in-one content factory
A platform for AI video production: avatars, voice-over, image generation. The whole content pipeline in one place.
Next.jsPythonRunwayElevenLabsReplicateGPU Inference