Applied General Intelligence, Inc.

Third Gen Foundational Intelligence
Austin

About Applied General Intelligence, Inc.

Applied General Intelligence is a new foundational AI system that solves the major issues around enterprises adapting AI today: quality, accuracy, privacy, transparency, attribution, and cost.

We are AI system born out of more than a decade of extensive research and computational breakthroughs. We are not an iterative or derivative improvement on current models but a whole new generation of AI. This gives us the ability to completely solve hallucinations, misalignment, transparency (fully explainable) and have significantly better accuracy while not requiring massive compute.
Specifically built for mission critical applications where quality and accuracy are paramount. Our technology allows companies to move beyond just a co--pilot/assistant to a full agent using just one model for the entire enterprise vs. having to spin up multiple verticalized models. This allows enterprises the ability to access, query, and synthesize all their unstructured data across their entire organization. This saves enterprises significant training costs, fine tuning, and expensive compute costs to train and run numerous models.

Tagged with

Team

Problem statement

For today's generative AI there are significant issues crossing the bridge from consumer to enterprise. This is due to enterprises having different objectives and demands: quality and accuracy, speed, cost efficiency, and above all, privacy for their most valuable asset, their data. Enterprises are faced with having to spin up multiple verticalized models vs. having just one enterprise-wide model. Additionally, these models still require human human oversight and quality review due to hallucinations and inaccuracies.

We solve these issues. It is one model made for the entire enterprise and made for mission critical applications. It ability to synthesize enormous amount of information (unstructured data) into insights

We have full Explainability, significantly better accuracy, alignment, no making things up and unlimited context window.

Traction information

Phase two pilot with Salesforce, Proof of concept with Oracle, LOI with prime defense contractor. On target to have a v1 demo March 2024

Milestones

November 2019

SQuAD Test

Using a very early version of our system we came in second on the Stanford Question and Answer Dataset (SQuAD) v. 1.1 test. We beat all the FAAG teams except one Google team by a fraction.

Updates

Lyndon Lay has visited this profile using a private link.
Added 3 months ago
Patrick has visited this profile using a private link.
Added 4 months ago
Feng Chi Wang has visited this profile using a private link.
Added 12 months ago
Will Baizer has visited this profile using a private link.
Added about 1 year ago
Nicole Bentz has visited this profile using a private link.
Added about 1 year ago
Nick Spiller has visited this profile using a private link.
Added over 1 year ago
Nick Spiller has visited this profile using a private link.
Added over 1 year ago
Feng Chi Wang has visited this profile using a private link.
Added over 1 year ago
Remi Donnelly has visited this profile using a private link.
Added over 1 year ago
Feng Chi Wang has visited this profile using a private link.
Added over 1 year ago

Network

Funding

Not raising capital right now

$600,000
committed
$500,000
goal
Total raised to date:$1,400,000