PRJCT LAZRUS  /  Model Intelligence

AI model intelligence, in real time.

Model-space intelligence and a repository build advisor.

NeoPlus monitors the entire AI model ecosystem — new releases, research breakthroughs, LLM benchmarks, and repo health — so you never miss what matters.

Get early access Explore features
29+ Models tracked
6 hrs HuggingFace poll interval
8 Benchmark sources
A–F Repo risk grading
Core platform

What NeoPlus does

Six integrated capabilities that give you a real-time operational picture of the AI landscape — from model releases to your own repo health.

Intelligence feed

Model Space Tracker

Polls HuggingFace every six hours across 10+ curated organizations — Google, Meta, Mistral, Qwen, DeepSeek, OpenAI, NVIDIA, and more. Tracks new models, updates, license changes, deprecations, major releases, and security advisories in a unified feed with source-specific filtering.

Research frontier

Research Intelligence

Monitors eight key ML research repositories and four arXiv categories for commits, experiments, releases, and papers. Detects experiment metrics in commits — val_bpb, loss, accuracy, perplexity — and surfaces breakthroughs automatically.

Benchmarks

LLM Leaderboard

Ten models against eight independent benchmarks — HumanEval, SWE-bench, MMLU, MATH, GSM8K, LiveCodeBench, GPQA, AIME — with category filtering, sortable columns, and an open-weight toggle.

Repo health

Build Mapper + Advisor

Scans your connected GitHub repositories to produce a canonical Build Map — detected models, frameworks, infrastructure, and risk flags with A–F grading. Generates ranked recommendations with implementation steps and file-level patch plans.

Model catalog

AI Model Explorer

29 models from 12 providers with full specifications — parameters, context windows, pricing, capabilities, API endpoints, and trending rankings. Search, filter by category or provider, and compare models side by side.

Alerting

Smart Notifications

Event-driven alerts for model updates, license changes, and security advisories, scoped to the repositories and providers you actually depend on. Ten YAML-based rules cover reproducibility, security, performance, and compliance.

How it works

Three layers of intelligence

NeoPlus combines real-time data ingestion, automated analysis, and actionable recommendations into a single mobile-first experience.

01

Ingest — continuous monitoring

Automated workers poll HuggingFace, GitHub research repositories, and arXiv on configurable intervals. Rate-limit aware, with graceful per-organization degradation and relevance filtering.

02

Analyze — structured intelligence

Raw signals are classified into event types, scored for relevance to your connected repositories, and enriched with benchmark data and experiment metric extraction.

03

Act — actionable output

Recommendations surface as prioritized cards, P0 through P3, with ordered steps and patch plans. Accept, dismiss, or apply — each with full status tracking.

Architecture

Built to last

Production-grade infrastructure designed for reliability, security, and real-time data at scale. Every layer is optimized for the developers who depend on it.

React Native + Expo Mobile-first client, TypeScript throughout, Redux Toolkit state
Node.js / Express + Prisma PostgreSQL with background workers and Redis cache
Firebase Auth + AES-256-GCM Authentication and encryption of stored access tokens
Octokit + arXiv + HuggingFace Repository scanning and upstream ecosystem ingestion
build_mapper.js 
class BuildMapper {
    /** Scan repos for ML best practices. */

    async scanRepository(repo) {
        // Detect models & frameworks
        const buildMap = await this.analyze(repo);

        // Apply rule engine
        const grade = this.ruleEngine.evaluate(
            buildMap, { rules: 'all' }
        );

        // Generate recommendations
        return this.advisor.recommend(grade);
    }
}

The AI landscape moves fast. You should move faster.

NeoPlus is currently in development. Join the waitlist to get early access.

Get early access Read the product page