Document parsing for dense files
Extracts tables, diagrams, reading order, and multi-column layouts from complex PDFs so downstream workflows can work with cleaner source material.
Lettria
ПодтвердитьLettria is an AI-powered document intelligence platform for regulated industries, turning unstructured documents into structured, traceable knowledge.
LettriaOverview
Lettria is an AI-powered document intelligence platform for regulated industries. Its stated purpose is to transform unstructured documents into structured knowledge that can be queried, traced, and reviewed with source context intact.
The product centers on four modules: Document Parsing, Ontology Building, Text to Graph, and GraphRAG. The site positions these tools as a pipeline for turning complex files into auditable knowledge graphs and context-rich answers for teams that need reliability, interpretation, and traceability.
The homepage highlights use in healthcare, finance, legal, and engineering, with examples such as evidence-based research, financial disclosure analysis, contract and regulation review, and technical troubleshooting. The overall workflow is built around extracting structure from documents first, then using that structure to support downstream analysis and decision-making.
Lettria also describes a native graph infrastructure called Perseus, which can connect documents, SQL databases, or APIs into a single graph and preserve traceability back to the original source. In the source material, the product is presented as a fit for organizations that work with high-risk or information-dense documents and need outputs that are understandable and reviewable.
Core capabilities
Ontology building
Automatically generates domain-specific ontologies from documents, reducing manual mapping and letting the structure evolve as the domain grows.
Text to Graph
Converts text into a knowledge graph with entities, relations, and constraints for structured analysis and reuse.
GraphRAG
Combines graph retrieval with reasoning to produce interpretable answers grounded in the underlying knowledge graph.
Dataset management
Provides a dataset manager for organizing parsed sources and reusing them across teams and projects.
Auditable graph infrastructure
Supports a graph layer, called Perseus, that can unify PDFs, SQL databases, or APIs into a production graph with source traceability.
Common use cases
Evidence-based research
Medical and scientific teams can query clinical trials, publications, and internal documents with answers grounded in explicit sources, evidence, and context for review and validation.
Financial disclosure analysis
Financial teams can extract and organize data from disclosures so AI systems can answer questions with traceable and auditable support.
Regulation and contract review
Legal teams can review regulations, contracts, and policy documents while keeping control over definitions, interpretation, and source grounding.
Technical document analysis
Engineering teams can analyze technical documents and troubleshoot issues using verifiable, step-by-step information instead of loosely connected summaries.
Knowledge graph preparation
Organizations building knowledge layers can turn parsed documents into reusable datasets and production graphs that support later GraphRAG workflows.
Pros and Cons
Pros
- Combines parsing, ontology building, graph creation, and GraphRAG in one workflow.
- Emphasizes traceability back to source documents and interpretable outputs.
- Supports several document types, including PDFs, images, spreadsheets, audio, and JSON.
- Provides dataset management for reuse across teams and projects.
- Positions the product for regulated or high-risk document workflows where review and validation matter.
Cons
- The public pricing page in the provided text does not show actual prices or plan details.
- The collected sources do not provide a full list of integrations or deployment options.
- The security page text is partial, so compliance specifics should be verified directly before procurement.
FAQ
What is Lettria used for?
Lettria is a document intelligence platform built to extract structure and verified knowledge from complex, high-risk documents. The source pages describe it as suitable for regulated industries and for teams that need traceable, context-rich answers from their own documents.
How does Lettria work?
The site says Lettria works through four core modules: Document Parsing, Ontology Building, Text to Graph, and GraphRAG. Together they turn unstructured documents into structured knowledge and interpretable outputs.
What kinds of files can it handle?
The document parsing page says Lettria can process text files, spreadsheets, PDFs, images, audio, and JSON. It also describes outputs such as full extracted text, document chunks, and metadata.
How is Lettria priced?
The pricing page does not show plan prices or limits in the provided text. It does show calls to action to request a demo or book a call, so pricing appears to be handled through direct sales contact on the public site.
Is Lettria secure enough for regulated work?
The source material mentions secure processing, traceability back to source documents, and a private environment for document parsing. It does not provide a full compliance certification list in the collected text, so any deployment or compliance requirements should be confirmed directly with the vendor.
Quick Facts
- Category
- Document intelligence / knowledge graph platform
- Primary users
- Teams in healthcare, finance, legal, and engineering
- Core workflow
- Parse documents, build ontologies, convert text to graphs, and query with GraphRAG
- Input types mentioned
- Text files, spreadsheets, PDFs, images, audio, and JSON
- Source domain
- lettria.com
- Pricing
- Demo or call requested on the public site; no prices shown in the collected text