ML Developer for Snippets Team (Core & Agentic Search)

Company: Alice AI Search Location: Moscow

You will be improving snippets — website cards in search results: launching new generative LLM-based snippets, turning ideas into tasks at the intersection of NLP and classic ML, inventing features and targets, and implementing solutions in C++ runtime. We expect you to know Python and C++, and to have experience solving NLP tasks.

A good search engine should not just find relevant documents, but also instantly answer the user's query directly in the search results. Our team is responsible for the presentation of entire documents — we create snippets. However, in 2026, the landscape has fundamentally changed. Search is no longer solely for humans. The agentic search scenario is rapidly growing, where the search engine acts as a key tool and external memory for autonomous AI agents.

High-quality snippets are critically important for agents: they allow for the instant extraction of the essence of a document directly from the search results, eliminating the need to load entire large web pages into the context window (which saves context, reduces latency, and lowers cost). We compete in the global market with products like Exa, Parallel, Tavily, and are building the next-generation search infrastructure. The snippets team is still small, but right now, in the era of agents, we are rethinking snippets and actively experimenting with new formats and technologies. Snippets represent an extractive and abstractive summarization task that we solve at the intersection of advanced NLP technologies.

Within the team, there is a startup spirit: each developer works on their own unique project and has complete freedom for experimentation. If you are ready to tackle full-cycle tasks — from forming hypotheses to rolling out improvements into production, we will be happy to see you on our team!