LUSI LABS

Carlo Ruiz, César Trejo
Last update: 3/17/2024

Request a project

Abstract

We build futuristic web applications with the help of LUSI ™, our revolutionary business AI. Our mission is to convert your idea into reality.

We have 6+ years of experience consulting and developing software for Silicon Valley companies such as thrivemetrics.com and finaleinventory.com.

Philosophy

We strive to understand your business goals and design technical solutions to meet them in minimal time. We take cybersecurity very seriously and work to ensure the secrety and integrity of the Software.

Projects

  1. Totaldevs. A platform for an exclusive community of professional latin american developers.
  2. Xavier. Discover the hottest upcoming products using artificial intelligence.
  3. Cursive. (Chrome extension) Measures and analyzes your writing process.

Open Source

  1. FastApy. Scalable python backend with job scheduling using rq and FastAPI.
  2. Pydyode. Run scientific python in the browser.

Research

The purpose of software is automation and generalization. We believe software is the tool that solves decision-making under uncertainty and constrained resources. Machines should be at the service of humans. Therefore, we focus our attention and efforts into developing safe and efficient algorithms for the benefit of humanity.
LUSI is a computable approximation of AIXI
(Hutter) An optimal agent (AIXI) is an agent that interacts with an environment in cycles k=1,2,...,m. In cycle k, AIXI takes action aka_k based on past perceptions o1r1…ok−1rk−1o_1 r_1\ldots o_{k-1} r_{k-1} as defined below. Thereafter, the environment provides a (regular) observation oko_k (e.g. a camera image) to AIXI and a real-valued reward rkr_k.
ak:=amax⁡ak∑okrk..max⁡am∑omrm(rk+..rm)∑q2−ℓ(q)a_k := \mathrm{a}\max_{a_k} \sum_{o_k r_k}.. \max_{a_m} \sum_{o_m r_m} (r_k + .. r_m) \sum_q 2^{-\ell (q)}
q=q:U(q,a1…am)=o1r1…omrmq = {q:U(q, a_1\ldots a_m)=o_1 r_1\ldots o_m r_m}

Contact

Send us an email.

Privacy policy

Terms of Service