There’s alpha hiding in your data. Lettria helps find it.
Between specialized jargon, number-heavy charts, and the need to find the smallest edge, applying standard AI solutions to financial information is a challenge. That’s why our teams have worked to develop new technologies that can properly read and interpret your company’s documents and data.
Three key principles
Accurate processing
Our NLP experts have developed AI technology that is better able to read and understand information and contextual relationships in your documents.
Transparent and verifiable
Every answer is accompanied by the relevant graphs and snippets that can be easily checked to ensure accuracy and peace of mind.
Ease of use
The Knowledge Studio is designed to be used by anyone in your company, regardless of their technical skills.

Reading financial charts
There’s a reason why there are so many examples of LLMs doing poorly on math. Standard AI solutions have an extremely hard time reading charts and numbers, as well as conducting multi-hop reasoning. Lettria’s parsing technology takes a giant leap forward, ensuring the greater accuracy and verifiability you need to effectively use AI in finance.
Relationships lead to insights
By creating knowledge graphs and nodes that understand complex relationships, Lettria’s AI technology can find the subtle signals hiding in your data. Whether it’s a new arbitrage or a trend change, you can investigate new opportunities and see the web of connections uncovered by the Knowledge Studio.


Personalized advice
Easily create specific use cases to adapt to the different needs of your clients and save time in the process. Whether you want to take risk profiles, regulatory requirements, or any other factor into account, it’s easy to create new information pools to query. And of course, every answer is transparent and verifiable thanks to the provided context.
Want to see Lettria’s Knowledge Studio in action?
If your company could use transparent, verifiable AI to improve performance, reach out to set up your personalized demo. And if you have other questions, such as about implementation, pricing, or more details on our GraphRAG technology, please don’t hesitate to ask.

Frequently Asked Questions
Our pilot program has our team of experts and engineers working directly with you to get the technology up and running, from document input all the way through to setting the Lettria platform up within your information system. And once it’s up and running, the platform has been designed to be intuitive and easy to use, no technical skills required. We’re also available to help whenever you need it, from troubleshooting to developing a new application within your business.
Lettria’s GraphRAG technology is able to virtually eliminate hallucinations in many contexts and can consistently best VectorRAG results by up to 30%. You can check out our full benchmarking between the two here.
Every time you query your documents, the Lettria platform shows you not only the answer, but also the snippets and relationships that led it to provide that answer. By having all of this information side by side, you can quickly check to make sure the machine provided the right answer – no leap of faith required.
It all comes down to a parsing problem: machines have to have the right information in the right format in order to provide accurate answers, and standard AI technology doesn’t give it to them. Lettria’s expertise in natural language processing gives us a step up when it comes to developing better ways for machines to interpret complex, unstructured data. We then take things further by combining the speed of vector-based RAG with the accuracy of knowledge graphs and GraphRAG. The result is industry-leading technology that can properly understand and query all kinds of documents, including those featuring lots of numbers, charts, and tables.