Metatext

Main Features:

1. Text Classification: Supports binary, multi-class, multi-label, sentiment, topic, and intents classification to categorize any text or document.
2. Text Extraction: Extract key information, entity recognition, and keywords to extract, normalize, and structure key information from any text or document.
3. Text Generation: Generative AI applications, RAG, summarization, LLM fine-tuning, Q&A, and chatbots to generate text from scratch or fine-tune LLMs to your domain.

Key Features & Core Advantages:

• No-code platform: No coding or ML/NLP expertise required. Just explain what you need to the model and let it learn from your data and expertise.
• AutoNLP engine: Put your data and expertise into the platform and let it do the rest. The AutoNLP engine will train and fine-tune the best model for your use case.
• LLM Distillation methodology: Refine your data into private and scalable custom models in minutes, combining the power of LLMs into an iterative process to efficiently train smaller models x10 better with less data and time.
• Trustful Evaluation: Helps you understand how your model is performing and how to improve it. Trustful models are key to success.
• Fast Deployment & Integration: Deploy your model in a few clicks and integrate with systems via API, Zapier, Google Sheets, Docker, AWS, Hugging Face, and more.
• No black-boxes and no vendor lock-in: Build and manage custom NLP models without black-boxes, vendor lock-in, or data privacy issues.

Typical Use Cases:

• Customer Support: Quickly categorize and respond to customer tickets, leaving more time for complex issues.
• Customer Insights: Automatically categorize reviews by topic, issue, sentiment, and engagement.
• Content Moderation: Automatically detect and remove inappropriate, offensive, or harmful content.
• Healthcare: Extract critical data from medical records, lab reports, and patient feedback.
• Finance: Analyze market sentiment, reports, and customer feedback.
• Human Resources (HR): Identify the best-fit candidates by matching skills, experience, and qualifications to job requirements.

Target Users: Developers, business analysts, data scientists, and anyone seeking to rapidly and efficiently train and deploy NLP models for text classification needs.

Pricing & Plans: Uses a Credits system for training models or making predictions (inferences) via API or inference preview. Offers three tiers:

1. Starter (Free): 500 credits/month, up to 5 projects, up to 1,000 records/project, up to 10 model versions/project, unlimited labels, public visibility, low GPU capacity, standard support.
2. Pro ($185/month): Everything in Starter plus 80,000 credits/month, unlimited projects, up to 10k records/project, unlimited model versions, visibility control, superior GPU capacity, priority support.
3. Enterprise (Custom): Everything in PRO plus custom pricing and resources, on-premise, dedicated AI engineer. Discounts available for academic research.

FAQ Info: Supports multiple languages; allows uploading custom datasets (CSV, TXT, PDF, DOCX, etc.) and importing via API; unlimited labels for text classification; trained models cannot be downloaded directly but can be integrated via APIs; Metatext is designed for creating and deploying custom NLP models, unlike ChatGPT and GPT which are general-purpose language models focused on generating human-like text and conversational AI.