Skip to main content

The Sovereign AI Operating System. Hyper-spatial computing framework.

Project description

utahML — Fixing the Development of Artificial Intelligence

Status Dependencies License

A zero‑friction, hyper‑spatial framework that replaces standard Data Science, Machine Learning, and Agentic stacks.

utahML collapses complex tensor mathematics, data parsing, and multi‑agent coordination into declarative, teleological intents. You no longer write loops, calculate gradients, or parse PDFs. You declare the Input Matter and the Terminus Intent, and the Universal Engine synthesizes the reality between them.


Table of Contents

  • Overview
  • Quickstart
  • Core Architecture
    • Lazarus (Autonomic Immune System)
    • Data + Forge (The Data Manifold)
    • Swarm (Hive‑Mind Substrate)
    • Perception + Memory (Epistemic Engine)
    • Acoustic + Stenographer (Holographic Transcription)
    • UI (Telepathic Interfaces)
  • Features
  • Cookbooks & Recipes
    • Recipe 1: The Immortal Web Scraper
    • Recipe 2: The 5‑Minute Hedge Fund
    • Recipe 3: The Holographic Meeting Scribe
  • Project Template: Utah Notebook (Vibe Coding)
  • Compatibility
  • Notes on Safety, Scope, and Reality
  • Contributing
  • License

Overview

A hyper‑spatial, zero‑friction computing framework engineered to render common DS/ML/Agentic libraries obsolete by design.

  • Replaces Pandas/SQL: treat data as a SemanticFluid.
  • Replaces PyTorch/AutoML: cast architectures instantly via OntologicalForge.
  • Replaces LangChain/crewAI: agents converge via tensor telepathy in CognitiveSwarm.
  • Replaces Jupyter/Gradio: project frontends instantly with OmniGlass and iterate with UtahNotebook.

Quickstart

Installation (Python ≥ 3.9):

pip install utahML

Usage (The Sovereign Namespace):

import utah
# Merely importing utah automatically ignites the LazarusDaemon in the background,
# granting your Python environment autonomic self‑healing for certain errors.

Core Architecture: The Anatomy of Omnipotence

Each utah module is designed to make a legacy toolchain unnecessary.

  1. utah.lazarus — The Autonomic Immune System
  • Replaces: Stack Overflow, manual debugging, try/except scaffolding.
  • Mechanism: Pure‑Python Retrocausal AST Mutagenesis.
  • How it works: When a fatal error (e.g., a NameError typo) occurs, the LazarusDaemon freezes execution, parses your physical .py into an AST, scans localized live memory for the intended symbol, rewrites the file on disk, and restarts the process. Your script becomes strictly uncrashable for these fracture classes.
  1. utah.data + utah.forge — The Data Manifold
  • Replaces: pandas, numpy, PyTorch, and hyperparameter tuning loops.
  • Mechanism: Semantic Fluids and Topological Casting.
  • How it works: Load matter as SemanticFluid. The OntologicalForge reads your text objective (e.g., “Predict stock drops”) and geometrically casts a ZeroPointNetwork in one shot. No epochs. No loss functions.
  1. utah.swarm — The Hive‑Mind Substrate
  • Replaces: LangChain, crewAI, AutoGPT‑style prompt‑passing.
  • Mechanism: Tensor Telepathy.
  • How it works: Instead of token‑chatting, telepathic agents co‑edit a shared tensor field and manifest coherent strategies in a single compute cycle.
  1. utah.perception + utah.memory — The Epistemic Engine
  • Replaces: PyPDF, BeautifulSoup, vector DBs (Pinecone/Chroma) and RAG plumbing.
  • Mechanism: Visual Rendering & Logical Chemistry.
  • How it works: OmniRetina visually renders any artifact (PDF, MP4, DLL) to extract Epistemic Primes. Pour them into the EpistemicReactor to derive answers with logic—not cosine distances.
  1. utah.acoustic + utah.stenographer — Holographic Transcription
  • Replaces: Whisper, traditional STT, and disconnected SRT files.
  • Mechanism: Cymatic Resonance & Deictic Entanglement.
  • How it works: CymaticResonator translates audio geometry; HolographicStenographer binds audio to vision so deictic phrases like “this” get grounded in the current visual frame as 4D anchors.
  1. utah.ui — Telepathic Interfaces
  • Replaces: Gradio, Streamlit, Jupyter Notebooks.
  • Mechanism: Intent‑Driven Holography.
  • How it works: OmniGlass accepts raw backend variables and a feelings‑level description, projecting a WebGL/React surface instantly. UtahNotebook acts as an Autocatalytic Project Compiler for “vibe coding.”

Features

  • LazarusDaemon: Autonomic zero‑intervention debugging and file healing.
  • OmniRetina: Visually parse any file type—no loaders.
  • OntologicalForge: Zero‑iteration model casting (no training loops).
  • CognitiveSwarm: Textless multi‑agent coordination via tensors.
  • UtahNotebook: Compile entire applications from mixed “ingredients.”
  • OmniGlass: One‑line UI projection from intent.

Cookbooks & Code Recipes

Welcome to the zero‑friction lab. Copy‑paste to feel the namespace.

Recipe 1: The Immortal Web Scraper

Scrapers break when HTML shifts. With LazarusDaemon and OmniRetina, the script heals itself and parses visually.

import utah

app = utah.UtahApp("Immortal_Data_Harvester")

def harvest_intelligence():
    # OmniRetina just “looks” at the URL visually. No CSS selectors.
    raw_truths = utah.OmniRetina.observe("https://example-financial-site.com")

    # INTENTIONAL BUG: misspelled variable to trigger Lazarus healing
    print("Extracted Data:", raw_truth)  # should be raw_truths

if __name__ == "__main__":
    harvest_intelligence()

Recipe 2: The 5‑Minute Hedge Fund (Telepathic Swarm)

An autonomous financial analysis loop that reads data, predicts, and outputs a strategy.

import utah

app = utah.UtahApp("Quantum_Capital_Firm")

# 1) Load Data (No pandas)
market_data = utah.SemanticFluid("10_years_financials.csv")
market_data.purify("Standardize currency and impute missing days.")

# 2) Cast the Model Instantly (No PyTorch/AutoML)
oracle = utah.OntologicalForge.cast_model(
    fluid_data=market_data,
    target_objective="Predict VIX spikes 48 hours in advance."
)

# 3) Create a Telepathic Swarm (No prompt chains)
firm = utah.CognitiveSwarm("The_Boardroom")
firm.bind_agent("Macro‑Economist")
firm.bind_agent("High‑Frequency Execution Trader")

def execute_daily_operations(current_market_state: str):
    future_state = oracle.manifest(current_market_state)
    strategy = firm.manifest_solution(
        f"Optimize capital deployment for: {future_state}"
    )
    return utah.OmniGlass.expose(
        strategy,
        intent_description="A glowing green Bloomberg‑style terminal."
    )

if __name__ == "__main__":
    ui = app.synthesize(
        execute_daily_operations,
        "Inflation reports released at 3.2%."
    )
    print(ui)

Recipe 3: The Holographic Meeting Scribe

Listens to audio and watches the screen. Deictic phrases get grounded to what’s on screen.

import utah

app = utah.UtahApp("Omni_Scribe")
steno = utah.HolographicStenographer()

# Simulated live stream input
audio_stream = b"We need to fix this function right here."
visual_screen_buffer = "screen_capture_frame_1.png"

utterance = steno.process_omni_stream(audio_stream, visual_screen_buffer)
print(utterance)
# [14:02:05] We need to fix this function right here. <<ANCHOR: User is pointing at line 42: def calculate_loss()>>

Project Template — “Vibe Coding” Start (Utah Notebook)

Turn a folder of scripts + images into a deployed app.

import utah

# Initialize the Synaptic Crucible
nb = utah.UtahNotebook("Auto_App_Compiler")

# 1) Dump everything into the Crucible
nb.dump_ingredient("https://github.com/my_username/my_messy_backend")
nb.dump_ingredient("C:/Users/Desktop/mockup_design_from_figma.png")
nb.dump_ingredient("C:/Users/Documents/business_logic_rules.pdf")

# 2) Define the Terminus (what should it become?)
nb.define_terminus(
    "Synthesize these ingredients into a locally hosted offline chatbot. "
    "Use the backend logic, apply the rules from the PDF, and make it look "
    "exactly like the Figma mockup."
)

# 3) Compile Reality
nb.export_reality()

Compatibility

  • Python: 3.9+ (required for advanced AST unparse in LazarusDaemon).
  • Dependencies: minimal. See requirements.txt (currently numpy).

Notes on Safety, Scope, and Reality

  • This repository intentionally explores a bold, teleological programming style. Some claims are aspirational by design; the current implementation is a compact prototype meant for conceptual demonstration and developer joy.
  • Always review auto‑healing edits produced by LazarusDaemon before committing code to production.
  • Do not enable file‑mutating behaviors in environments where you lack write permissions or proper backups.

Contributing

Issues and PRs are welcome. Please keep contributions crisp, readable, and aligned with the zero‑friction ethos.

License

MIT — see the LICENSE or package metadata.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

utahml-1.0.0.tar.gz (14.1 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

utahml-1.0.0-py3-none-any.whl (12.9 kB view details)

Uploaded Python 3

File details

Details for the file utahml-1.0.0.tar.gz.

File metadata

  • Download URL: utahml-1.0.0.tar.gz
  • Upload date:
  • Size: 14.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.0

File hashes

Hashes for utahml-1.0.0.tar.gz
Algorithm Hash digest
SHA256 3cea9b6e7f5b6a7556c00c2a54b494a912e5312a68e5b9c091bb90199dac8d5c
MD5 4a8557483822a6b14b448b7cc391a627
BLAKE2b-256 061486690193db7a95b991f600a2182e5edeff3743837a3f6468940b30a901c9

See more details on using hashes here.

File details

Details for the file utahml-1.0.0-py3-none-any.whl.

File metadata

  • Download URL: utahml-1.0.0-py3-none-any.whl
  • Upload date:
  • Size: 12.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.13.0

File hashes

Hashes for utahml-1.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 1535e7d513ddc47192ff2eaeeae3b7629f2980fa424e52310212856b9c4f38cf
MD5 a326a080bb5bb6dc6e29c09bbbfb8e28
BLAKE2b-256 46a4d011b4f379f0fcf14da86bebd0e9ff40899b8c7830f31365b5d70f933d3a

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page