EnergeticAI

EnergeticAI: Use open-source AI in your Node.js apps is a high-performance, open-source AI toolkit built on TensorFlow.js optimized for serverless environments. It focuses on fast cold-start, small module size, and pre-trained models to help developers integrate embeddings, classifiers, and other AI capabilities into Node.js applications with minimal setup and licensing suited for business use.

---

Overview

• Provides pre-trained embeddings and AI models for recommendations, semantic search, classification, and question-answering (planned).
• Built on TensorFlow.js with a focus on serverless-friendly characteristics:
• Fast cold-start
• Small module footprint
• Optimized performance (up to 67x faster in certain benchmarks vs TensorFlow.js)
• Simple installation and usage via NPM. Requires Node 18+ and is Apache 2.0 licensed.
• Open-source ecosystem with model sources and bundled embeddings for quick iteration.

How to Use EnergeticAI

1. Install the core package: npm install @energetic-ai/core.
2. Import and initialize a pre-trained model:

• Example:
• import { initModel, distance } from "@energetic-ai/embeddings";
• import { modelSource } from '@energetic-ai/model-embeddings-en';
• (async () => { const model = await initModel(modelSource); const [hello, world] = await model.embed(["hello", "world"]); console.log(distance(hello, world)); })();

1. Use embeddings, classifications, and other available components in your app.

Notes:

• EnergeticAI emphasizes fast startup and small bundle size, enabling efficient inference in serverless environments.
• The library is designed to be integrated directly into Node.js backends for tasks like recommendations, semantic search, and classification.

Models & Libraries

• Pre-trained embeddings for English language tasks (e.g., sentence embeddings).
• Classifiers for English (text categorization) with minimal training examples.
• Planned features include QA models for meaning-based answering.

Performance Highlights

• Cold-start speed and module size optimized for serverless deployments.
• Benchmark comparisons show substantial speed advantages over standard TensorFlow.js in specific scenarios.

Real-World Privacy & Licensing

• Open-source with business-friendly licensing (Apache 2.0).
• Licensing and dependencies may vary by component.

How It Works

• Load a pre-trained model from a bundled source.
• Generate embeddings, distances, and other AI outputs for input text.
• Use these outputs for downstream tasks like similarity search, recommendations, or classification.

Core Features

• Serverless-optimized TensorFlow.js-based AI toolkit for Node.js
• Pre-trained embeddings and classifiers (English-focused) with ready-to-use APIs
• Fast cold-start performance and small module footprint
• Easy installation and integration via NPM
• Business-friendly Apache 2.0 licensed with open-source model sources
• Benchmark-driven performance improvements over standard TensorFlow.js
• Planned QA and additional NLP capabilities