turbopuffer hit $100M run-rate in 19 months with <$1M invested and profitability! turbopufferdb 👏🏼 Their thesis: AI doesn’t just need bigger models, it needs cheaper memory. Turbopuffer built search infrastructure on object storage, instead of relying mainly on expensive memory and SSDs. Why it matters: Every AI product needs to retrieve context. But as context grows, infrastructure costs explode. Turbopuffer’s bet: store more in cheap object storage, not expensive memory. Their clients include Cursor, Notion, Linear, Anthropic, Atlassian, Ramp, Harvey, Legora, Atlassian, Cognition, Clay, Superhuman… New fastest growing startup? #turbopuffer #AIInfrastructure #VentureCapital #Startupgrowth