Senior AI/LLM Integration Engineer

SilentRoom is a professional digital environment for authors, built on the 'Silent AI' philosophy. It combines classic editor structures with the power of Gemini 3.1 Pro. - Application operates in four modes: BUILD, THINK, WRITE, CHECK. - Stack includes Next.js, Supabase, Tiptap, Google Vertex AI. - AI modes: Project, Global, Search. - Integration with leading models like Gemini, OpenAI, Anthropic.

Take full responsibility for all LLM integrations in SilentRoom across three modes: - Project (RAG) with stateful context and context caching. - Global with stateless context and fast responses. - Search with web search and proof links. - Connect and support integrations with leading models.

We are looking for a top-senior level professional with: - Experience with LLM API (Gemini, OpenAI, Anthropic) in production. - Knowledge of RAG, vector databases (Pinecone, Weaviate, pgvector), embeddings. - Proficiency in Python or Node.js backend (FastAPI, Express). - Understanding of prompt engineering, context management, async architecture. - Experience with context caching (Google Vertex AI is a plus).

Work format includes project work and subsequent support: - 100% remote, any location. - Communication in Russian; high-level technical English. - Flexible payment options in any currency, including rubles.