← back
The Adversarial Path to the Personal Assistant: Sumit Agarwal
Takeaway
True personal AI requires aggregating a user's full data footprint via adversarial ETL — generic LLMs without that data give Google-plus answers.
Summary
- Sumit Agarwal (Ario, $16M raised) builds an AI personal assistant whose foundation is 'adversarial ETL' — extracting user data (Google, Facebook, DoorDash, Strava, Instagram) from systems that don't want to share it.
- System generates auto-derived 'data portraits' (e.g. SLR photography, daycare patterns, child interests) plus an editable memory layer for explicit user-told facts.
- Demos contextually personalized queries: 'where's a sunny warm place I can travel to' returns Maui (Asian food), San Diego (Legoland for Lego-buying son), Miami — vs. CoPilot's generic Google-plus answer.
- Goal: give every user 'one hour a day back' via personal context aggregated from across their digital footprint.
personal-assistantragmemory
Original description
Despite their trillions of tokens of training data, GPT4 and others models can’t answer a single question about us as unique individuals. The personal AI assistant we all want surely has to know us as individuals, right? Today’s AI assistants either give us generic information, or require significant input from us before providing useful personalized assistance. Isn’t there a better path? Of course there is. In this talk, we’ll discuss our implementation of a personal, private, proactive AI assistant that begins its relationship with you by downloading your transaction logs from major B2C sites (think of Google, Facebook, Amazon, Doordash, Strava, Uber and dozens of others) to your local device. By building a RAG system using this data, our personal AI assistant becomes useful immediately, without requiring significant user input. Learn how a modest amount of very powerful personal data can unlock the power and potential of personal AI assistants. Recorded live in San Francisco at the AI Engineer World's Fair. See the full schedule of talks at https://www.ai.engineer/worldsfair/2024/schedule & join us at the AI Engineer World's Fair in 2025! Get your tickets today at https://ai.engineer/2025 About Sumit Sumit Agarwal is founder and CEO of Ario AI, which is developing a personal agent built on top of a person's entire online data history. Before Ario, Sumit spent 8 years building Shape Security, the leading security vendor for user data protection. F5 Networks purchased Shape for $1B in Jan 2020. Immediately prior to Shape, Sumit was a Deputy Assistant Secretary of Defense at the Pentagon where he worked on national security issues. Sumit served as a network warfare officer in the United States Air Force Guard and Reserve for 17 years. He started his career with stints at Trilogy, in venture capital, and as a founder of Quova. He also spent 6 years at Google from 2003 to 2009. Sumit holds a BS in Chemical Engineering from MIT.