Does this connect to real meters?

No. It is an end-to-end simulation with generated heartbeat pulses, a fleet simulator and an anomaly injector that let you watch Jev decisions and remediation on synthetic data.

What Jev primitives does it use?

Each escalated pulse is classified with a choice over fault types such as tamper, battery decay and RF attenuation, scored from Tier 0 to Tier 4 and checked with a probability question for whether field dispatch is warranted.

Where do the cost and latency numbers come from?

From the project's own model, not a deployment. The README puts Jev at $0.042 per million input tokens with about 40 tokens per evaluation against roughly 450 and $0.02 for a generative LLM, which it rolls up to $32.71 a month for a 10M-meter fleet against a $214,500 baseline, on 80-140 ms evaluations averaging about 114 ms.

What does the edge pre-filter do before Jev sees anything?

The gateway caches 92.4% of nominal pulses locally, defined as zero delta, nominal voltage and clear tamper flags, and escalates only the remaining 7.6%, which are anomalous, borderline or cyclic verification failures, to the Jev layer as a compact typed-primitive query.

What remediation does it trigger?

Three paths: automatic SIM quarantine and revenue-fraud flagging for a security event, automated truck-roll work orders such as #WO-89104 with technician assignment, ETA and resolution lifecycle, and cellular beam steering with local interference flagging for RF problems. Live sliders adjust the dispatch probability threshold, severity cutoffs and edge pre-filter rate.

How do I run and test it?

Clone the repo, run npm install on Node 18 or newer, then npm run dev and open http://localhost:5173/. npm test validates the Layer 2 deterministic pre-filter, the Layer 3 typed-primitive contract and the Layer 4 remediation logic, and npm run build plus npm run preview give you the production build. It is MIT licensed.

What does the dashboard actually render?

A 2.5D canvas engine drawing over 1,500 animated micro-nodes for the 10M-endpoint fleet at 60 FPS without external map API keys, spread across five metro districts with critical beacons strobing red and major anomalies amber. Beside it, a live telemetry feed shows the raw 64-byte pulse, the classification and confidence, the severity tier, the dispatch probability and the executed action.

CommunityJev Demos & Experiments

Jev-IOT

React dashboard and simulator for a 10-million smart-meter fleet that triages telemetry through a deterministic edge pre-filter and then Jev Choice, Score and probability primitives, triggering SIM quarantine, work orders or RF fixes at about $33 per month in its own model.

See highlights

Open github.com

Stars: 1
Language: TypeScript
License: MIT
Published by: Community
Last verified: Oct 2026

pjmenon45/Jev-IOT

User repository on GitHub

View on GitHub

In a 10-million smart meter deployment, using generative LLMs (like GPT-4 or Claude) is economically impossible and operationally impractical due to token generation latency. Enter Jev - zero output token fees and ultra-low input cost ($0.042/M tokens), you achieve massive scale at negligible compute expense.

Forks: 0
Last pushed: Updated Sep 2026

Repo stats from the GitHub API, cached Oct 2026.

About this project

An end-to-end, carrier-grade edge AI telemetry classification and autonomic remediation dashboard designed for Tier-1 utility and telecom providers managing 10M+ smart utility meters (gas/water/electric). Powered by Jev—a non-autoregressive, typed-primitive "System One" AI engine that slashes inference costs from $0.02/evaluation down to $0.042 per million input tokens with zero generative token bloat and sub-150ms parallel evaluation latency.

Quoted from the project's README. Read the full README.

Tagscommunity iot simulation dashboard

Works forJev for Classification Jev for Scoring Jev for Decision gating

Used inJev for Data & Analytics

Languages: TypeScript
Platforms: Browser
Capabilities: Open source

Highlights

README models 92.4% of nominal pulses filtered at the edge and 7.6% escalated to Jev.
Each anomalous pulse is batched into a Choice fault class, a Score severity tier and a dispatch probability question.
The 2.5D canvas renders over 1,500 animated nodes at 60 FPS without map API keys.
Vite + React app runs with npm install and npm run dev; npm test exercises the pipeline.

Watch out

MIT licensed. Everything runs on simulated telemetry; the fleet size, latency and cost figures come from the project's own model rather than a deployment.

Questions about this entry

Short answers drawn from the linked project or page.

More Jev resources like this

WorldMonitor

88.2k

Real-time geopolitical intelligence dashboard that uses TypeSafe Jev questions to score news headline severity into five threat levels and categorize events across 14 conflict, cyber and infrastructure domains, alongside news aggregation and a Country Instability Index.

Jev Integrations & Adapters#community#osint#geopolitics

Community

koala73/worldmonitorTypeScript

Cua

29.3k

The Cua computer-use platform documents a jev-use recipe where TypeSafe Jev picks bounded action IDs and Cua Driver executes clicks and fills in the browser.

Jev Demos & Experiments#community#browser#computer-use

Community

trycua/cuaRust

jev-plays

A Craftax (Crafter) survival harness where deterministic code lists every feasible action with its facts and Jev picks one Choice per step, optionally guided by an LLM-written objective and standing rules; three-seed ablations are included.

Jev Demos & Experiments#community#craftax#games

Community

mansicer/jev-playsPython

FlightBench

Fixed-wing landing simulator where a human can compete against Jev: four Jev Choice decisions map to aileron, elevator, rudder and throttle commands in JSBSim, run in real time. The benchmark also compares Jev, classical controllers and Codex models on landing score and API cost.

Jev Demos & Experiments#community#simulation#aviation

Community

AlperKartkaya/FlightBenchPython

THE HUNDRED EYES

Interactive media artwork where one post meets 100 fictional observers: for each observer the app asks Jev one Choice plus four Score questions (500 questions total), animates the eyes in Three.js and later exposes the full distribution of reactions in a post-analytics view.

Jev Demos & Experiments#community#interactive-art#threejs

Community

mintannn/THE-HUNDRED-EYESTypeScript

Field Advisory

An agricultural advisory demo that turns sensor readings, a forecast, and a crop stage into a next-action choice plus an urgency score, routing marginal readings to an agronomist instead of triggering an application.

Jev Demos & Experiments#community#python#agriculture

Community

From the Jev guides

Original write-ups on what this entry is for.

Per-Tick Decisions in Games

Sep 28, 2026

A game tick is a state plus a legal action list. How Clash Royale, Minecraft, and the earlier Mario, Pokémon, StarCraft, and drone loops ask Jev one step at a time.

Jev Demos & Experiments#gaming#agents#control

Read article

Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai