A 19-year-old nabs backing from Google execs for his AI memory startup, Supermemory (techcrunch.com)

Nineteen-year-old Dhravya Shah announced Supermemory, a universal memory API designed to give AI apps longer, multimodal context by extracting “memories” from unstructured data and building a knowledge graph. The startup—born from a week-long hack project and shepherded by Shah’s Cloudflare internship/advisors—can ingest files, docs, chats, emails, PDFs, app streams and video, surface insights across time (e.g., month-old journal entries or email threads), and connect to tools like Google Drive, Notion, plus a Chrome extension for quick notes. Shah says Supermemory’s core strength is turning varied, noisy inputs into searchable context for apps such as email clients, writing tools, video editors and robotics systems. Supermemory closed a $2.6M seed round led by Susa Ventures, Browder Capital and SF1.vc, with angel backing from Jeff Dean, Cloudflare CTO Dane Knecht, DeepMind’s Logan Kilpatrick, Sentry’s David Cramer and execs from OpenAI, Meta and Google. It already powers customers including Cluely, Montra, Scira, Rube and Rets. Technically significant for the AI community, Supermemory positions itself as a low-latency memory layer that complements LLM context windows—enabling persistent, multimodal state and faster retrieval for agentic and app-focused workflows amid a crowded field (Letta, Mem0, Memories.ai) where performance and integration breadth will decide adoption.