19-Year-Old Secures Google Backing for AI Memory Startup, Supermemory

The Evolution of AI Memory and the Rise of Supermemory

The capacity of Artificial Intelligence models to retain information, often referred to as their context window, has been steadily increasing. Despite this progress, challenges remain in enabling AI to maintain long-term memory across multiple interactions.

Dhravya Shah, a 19-year-old entrepreneur, is addressing these challenges with Supermemory, a novel memory solution designed specifically for AI applications.

From Bot Creator to AI Innovator

Originally hailing from Mumbai, India, Shah began his journey by developing consumer-focused bots and applications. He successfully sold a bot that enhanced tweet screenshots to Hypefury, a social media management platform.

The revenue generated from this sale allowed Shah to pursue his education in the U.S., choosing to attend Arizona State University instead of preparing for the entrance exam to the Indian Institute of Technology (IIT).

Following his relocation, Shah embarked on a personal challenge: to create a new project each week for 40 weeks. It was during this period that Supermemory, initially known as Any Context, was born. The initial iteration allowed users to interact with their saved Twitter bookmarks.

Supermemory: Extracting and Utilizing Context

The current iteration of the tool focuses on extracting valuable “memories” and insights from unstructured data sources.

This process empowers applications to gain a more comprehensive understanding of the context surrounding user interactions.

In 2024, Shah completed an internship at Cloudflare, concentrating on AI and infrastructure. He subsequently transitioned into a developer relations leadership role at the company.

During his time at Cloudflare, advisors, including CTO Dane Knecht, encouraged him to develop Supermemory into a fully-fledged product.

This led to Shah’s decision to dedicate himself full-time to building Supermemory this year.

A Universal Memory API for AI

Supermemory is now positioned as a universal memory API for AI applications. It constructs a knowledge graph from the data it processes, tailoring the context to individual users.

For example, the solution can facilitate querying data from months prior within a writing or journaling application, or enable efficient searches within an email client.

Its ability to handle multimodal inputs also extends its utility to video editors, allowing them to retrieve relevant assets from a library based on specific prompts.

The startup is capable of processing diverse data types, including files, documents, chats, projects, emails, PDFs, and real-time app data streams.

Users can leverage its chatbot and notetaker features to add memories in text format, incorporate files or links, and connect to popular platforms like Google Drive, OneDrive, and Notion.

A Chrome extension further simplifies the process of adding notes directly from websites.

“Our primary strength lies in our ability to distill insights from any form of unstructured data, thereby providing applications with a richer understanding of their users,” Shah explained. “Our solution’s compatibility with multimodal data makes it applicable to a wide range of AI applications, from email clients to video editors.”

Securing Seed Funding and Industry Support

Supermemory has successfully raised $2.6 million in seed funding, led by Susa Ventures, Browder Capital, and SF1.vc.

The investment round also included contributions from prominent individual investors, such as Cloudflare’s Knecht, Google AI chief Jeff Dean, DeepMind product manager Logan Kilpatrick, Sentry founder David Cramer, and executives from OpenAI, Meta, and Google.

Shah mentioned that Y Combinator had also expressed interest in his project, but the timing was not ideal due to existing investor commitments.

Joshua Browder, founder and CEO of DoNotPay and managing partner of Browder Capital, was particularly impressed by Shah’s rapid execution and innovative spirit.

“I connected with Dhravya on X, and his ability to quickly develop and deploy solutions immediately caught my attention, prompting me to invest,” Browder stated.

Existing Customers and Competitive Landscape

Supermemory currently serves a growing number of customers, including Cluely, a desktop assistant backed by a16z; Montra, an AI video editor; Scira, an AI search engine; Composio’s Rube, a multi-MCP tool; and Rets, a real estate startup.

The company is also collaborating with a robotics firm to enable robots to retain and utilize visual memories.

While the application offers a platform for developers to explore the tool’s capabilities, it also serves as a valuable resource for understanding its potential applications.

The AI memory space is becoming increasingly competitive, with startups like Letta and Mem0 (where Shah previously worked) developing memory layers for AI agents.

Susa Ventures, an investor in Supermemory, has also invested in Memories.ai, alongside Samsung, which focuses on extracting insights from extensive video footage.

Shah believes that Supermemory will differentiate itself through its lower latency and superior performance.

“As AI continues to evolve, the need for a robust memory layer will only intensify,” Browder added. “Supermemory’s solution delivers high performance while ensuring rapid access to relevant context.”