Lettria

What is Lettria?

Lettria is an NLP platform built for software developers and knowledge workers. It combines Large Language Models (LLMs) with symbolic AI to overcome common challenges in extracting knowledge from text. The platform turns messy, unstructured documents into dependable information by understanding the connections between pieces of data within text. It focuses on sustainability, efficient resource use, explainability, data privacy, and transparency.

Lettria’s platform is built on three main principles:

1. Focused models: Smaller, specialized models for specific tasks, built using graph-based techniques.
2. Accessible tools: High-powered text processing and analysis available through a no-code platform.
3. Advanced technology: NLP and cloud computing used to analyze text quickly, accurately, and at scale, supporting customization.

The goal is to overcome the limitations of LLMs and reduce the effort required for NLP tasks while improving the odds of success. Lettria draws on a database of over 800,000 words, a detailed ontology, methods for handling words with multiple meanings, proprietary models, and user feedback to support its NLP projects. The company aims to make language models and NLP accessible to a wide range of users while encouraging collaboration and innovation in the field.

Who created Lettria?

Lettria was founded by Charles Borderie, Marian Szczesniak, and Victor de La Salmonière, who share backgrounds in entrepreneurship, AI development, and technology.

• Charles Borderie is a serial entrepreneur who has consulted in both Europe and the Americas.
• Marian Szczesniak is an AI expert and developer specializing in NLP applications.
• Victor de La Salmonière is an engineer with experience in DevOps, data science, and development.

The company’s focus is using natural language processing to help organizations extract useful insights from their data, while making technologies like language models and NLP accessible to a broad audience.

What is Lettria used for?

• Cleaning up text
• Mining text for insights
• Classifying text
• Crafting prompts
• Improving ontologies
• Turning text into graphs
• Creating private GPT instances
• Using GraphRAG (Retrieval Augmented Generation with Graphs)

Who is Lettria for?

• Software developers
• Knowledge workers

How to use Lettria?

Start on Lettria’s platform, which offers features for cleaning text, mining it for information, classifying it, and prompt engineering. A good entry point is Lettria Knowledge Studio, where you can take unstructured documents and turn them into organized, valuable information.

Many users choose Lettria to keep control of their data and to get better results than other NLP solutions, using its suite of APIs. By combining linguistic databases, LLMs, and open-source technologies, Lettria aims to change how text is processed across a range of industries.