Loading organizations...
§ Private Profile · San Francisco, CA, USA
Supermemory is a technology company.
Supermemory develops context infrastructure designed to provide artificial intelligence agents with advanced memory capabilities. The company offers a universal memory API that integrates deep user profiling, a custom vector graph engine for dynamic knowledge representation, and sophisticated retrieval mechanisms. This comprehensive approach allows AI systems to not only recall information but also to understand intent, preferences, and complex contextual relationships across diverse data formats.
The company was founded by Dhravya Shah, who initially created an open-source "second brain" application. Shah's foundational insight stemmed from the recognition that AI's effectiveness is intrinsically linked to its ability to retain and process context beyond immediate interactions. This led to the development of a more robust, evolving knowledge system rather than a static data store.
Supermemory serves both developers and teams through its API, which is utilized for advanced retrieval, augmentation, and extraction within AI applications, and individual users via its personal Supermemory app and plugins. The company's vision is to establish the definitive memory layer for AI agents, enabling them to achieve deeper understanding and deliver more personalized, intelligent experiences by ensuring AI systems have a complete and evolving knowledge base.
Supermemory has raised $29.0M across 2 funding rounds.
Supermemory has raised $29.0M in total across 2 funding rounds.
Supermemory has raised $29.0M across 2 funding rounds. Most recently, it raised $26.0M Seed in October 2025.
| Date | Round | Lead Investors | Other Investors | Status |
|---|---|---|---|---|
| Oct 6, 2025 | $26M Seed | Browder Capital, Scientifica Venture Capital, Susa Ventures | Dane Knecht, David Cramer, Jeff Dean, Logan Kilpatrick | Announced |
| Oct 1, 2025 | $3M Seed | — | Banana Capital, Better Tomorrow Ventures, Browder Capital, Davidovs VC, Foundry Group, Kepler Operator’s Fund, LAUNCH, Resolute Ventures, ROO Partners, Sarah Smith Fund, Stellar Capital, Todd And Rahul's Angel Fund, Trust Fund, Bobby Goodlatte, David J. Phillips, Diego Oppenheimer, Roger Dickey | Announced |
Supermemory has raised $29.0M in total across 2 funding rounds.
Supermemory's investors include Browder Capital, Scientifica Venture Capital, Susa Ventures, Dane Knecht, David Cramer, Jeff Dean, Logan Kilpatrick, Banana Capital, Better Tomorrow Ventures, Davidovs VC, Foundry Group, Kepler Operator’s Fund.
Supermemory is a San Francisco-based AI startup building a universal memory API and infrastructure layer for AI applications, enabling persistent, scalable memory that adapts like the human brain.[1][2][3] Founded by 20-year-old Dhravya Shah, it solves AI's core challenge of long-term context retention by constructing dynamic knowledge graphs from user data streams, supporting multimodal inputs across apps like AI assistants, video editors, healthcare tools, and enterprise search.[1][2][3] It serves developers, enterprises, and builders—powering hundreds of customers processing over 5 billion tokens daily—with use cases in education, healthcare, legal compliance, and robotics, recently achieving state-of-the-art performance on LongMemEval benchmarks for agent memory.[1][3][4] After raising $3 million, Supermemory has shifted from an open-source side project to a commercial product with strong growth, backed by investors including Google's AI executives and Cloudflare advisors.[2][3]
Supermemory began as an open-source bookmarking and notetaking tool that Dhravya Shah, a Mumbai-born prodigy, built as a side project in his dorm at age 18, around 2023.[2][3] Shah, now 20 and a college dropout, interned at Cloudflare in 2024 on AI and infrastructure projects, rising to developer relations lead before advisors like Cloudflare CTO Dane Knecht urged him to commercialize it.[2] Interest from enterprises willing to pay for implementations prompted Shah to drop out, relocate to San Francisco as a solo founder, and pivot to a full memory API—declaring it his "life's work."[3] Early traction came from diverse AI apps, leading to $3 million in funding from top AI investors and rapid scaling to enterprise clients handling billions of tokens weekly.[2][3]
Supermemory rides the agentic AI trend, where models must maintain long-term coherence for real-world apps amid exploding data volumes—addressing "one of AI’s hardest challenges" as LLMs shift from stateless chat to persistent, personalized agents.[1][3][4] Timing is ideal post-2024 AI infrastructure advances (e.g., Shah's Cloudflare experience), with market forces like multimodal data growth and enterprise demand for compliant memory (healthcare, legal) fueling adoption.[1][2] It influences the ecosystem by providing a foundational layer—much like vector databases did for RAG—enabling builders in education, robotics, and search to create context-aware tools, potentially standardizing memory for the next wave of AI infrastructure.[3][4]
Supermemory is poised to dominate AI memory as agentic systems proliferate, with its SOTA benchmarks and enterprise scale positioning it for rapid expansion into more verticals like robotics visuals and procurement AI.[1][3][4] Trends like multimodal agents and edge inference will amplify its edge, especially with backing from AI heavyweights; expect partnerships, further funding, and open-source evolutions to cement it as infrastructure plumbing. As Shah scales from solo founder to team, Supermemory could redefine how AI "remembers," turning ephemeral interactions into enduring intelligence—echoing its dorm-room origins into Silicon Valley disruption.[2][3]