At just 19 years old, Dhravya Shah has raised $2.6 million in seed funding for his AI startup Supermemory, which aims to solve one of artificial intelligence’s biggest limitations: long-term memory.
The round was led by Susa Ventures, Browder Capital, and SF1.vc, with participation from prominent angels including Jeff Dean (Chief Scientist, Google), Dane Knecht (CTO, Cloudflare), Logan Kilpatrick (Product Manager, DeepMind), David Cramer (Founder, Sentry), and executives from OpenAI, Meta, and Google.
From Mumbai to building a global AI infrastructure
Originally from Mumbai, India, Dhravya Shah began building consumer-facing bots and apps as a teenager. One of his early projects — a tool that turned tweets into screenshot-ready images — was acquired by Hypefury, marking his first successful exit before college.
Shah was preparing for India’s elite IIT entrance exams when the sale inspired him to take a different path. Using the proceeds, he moved to the United States to study at Arizona State University, where he began experimenting with AI projects.
During a self-imposed challenge to “build something new every week” for 40 weeks, he created the first version of Supermemory — then called Any Context — as a side project that let users chat with their Twitter bookmarks.
By 2024, after an internship and a stint leading developer relations at Cloudflare, Shah was encouraged by mentors — including Knecht — to turn Supermemory into a full-fledged company.
Advertisement
Giving AI a memory that lasts
Today, Supermemory is described as a universal memory API for AI apps, designed to give artificial intelligence systems persistent, context-aware memory.
The platform processes unstructured data — from documents, chats, projects, PDFs, and emails — and builds a knowledge graph that allows apps to retain and recall relevant context over time.
“Our core strength is extracting insights from any kind of unstructured data and giving AI apps more context about users,” said Shah. “By working across multimodal inputs — text, images, video, and more — we make every AI tool smarter and more human-aware.”
Through integrations with Google Drive, OneDrive, Notion, and a Chrome extension, users can easily add “memories” from across their digital footprint. Developers, meanwhile, can use Supermemory’s API to embed memory capabilities into their own apps.
Early traction and customer base
Supermemory’s technology is already being used by Cluely (an a16z-backed desktop assistant), Montra (AI video editor), Scira (AI search), Composio’s Rube, and Rets (real estate startup). The company is also partnering with a robotics firm to help machines retain visual memory from their environments.
While developers are the initial target users, Shah sees Supermemory as a foundation for all future AI applications: “More and more AI products will need a memory layer — one that’s fast, contextually aware, and privacy-safe.”
Investor conviction in the next generation of AI infrastructure
Joshua Browder, founder and CEO of DoNotPay (and founder of Browder Capital), said he was drawn to Shah’s execution speed. “What struck me about Dhravya is how fast he builds and learns. That tenacity is exactly what this space needs,” Browder said.
Other investors agree that long-term AI memory could be the next major leap in model capability. Competing startups like Letta, Mem0, and Memories.ai are also exploring the space, but Shah believes Supermemory will stand out for its low latency and flexible multimodal architecture.
“Supermemory’s platform gives developers the ability to surface relevant context instantly — that’s what will make or break user experience in next-gen AI tools,” Browder added.
What’s next
With the new funding, Shah plans to expand the engineering team, improve Supermemory’s latency and scalability, and continue refining its multimodal data engine.
Reflecting on the journey, Shah said: “AI can already think — but it doesn’t yet remember. We’re building the infrastructure that changes that.”
About Supermemory
Founded in 2024 by Dhravya Shah, Supermemory is an AI infrastructure startup developing a universal memory API for AI apps. Its platform extracts, organises, and recalls insights from unstructured data, allowing applications to build long-term contextual understanding. The company is backed by Susa Ventures, Browder Capital, SF1.vc, and leading AI researchers from Google, Cloudflare, and OpenAI.
