Заявка Compresr в YC

Compresr

LLM context compression for better accuracy

We built an API that compresses the input context for large language models without losing critical information. This lets developers using agents or retrieval-augmented generation cut token costs and improve output accuracy. Our system identifies what matters in prompts and shrinks the rest, making LLMs more efficient in real use cases.

We are currently remote; the company is not fixed to one location. After YC, we would likely remain flexible or move near Silicon Valley to leverage the YC network.

We chose to stay remote initially to tap talent anywhere and maintain speed. Post-YC, proximity to YC and Silicon Valley would open doors for partnerships, hiring, and investment - it's where a lot of LLM infrastructure innovation is happening now.