Natural Language Processing (NLP) at Lingua Custodia

🔎 Inside Lingua Custodia’s lab!

Our head of research Raheel will answer 3 questions that will help you understand the LAB, its latest developments and a forecast on the future of NLP at Lingua!

▶ What do you do at the LAB? How does it work?:

At Lingua Custodia Lab, we have a team of 5 researchers, 3 researchers with PhDs in machine learning, one current PhD sudent in natural language processing and a master intern.

The responsibilities of the lab are 3 fold. First, we follow state-of-the-art in natural language processing and machine learning and try to bring the newest techniques to our production models.
Secondly, we conduct fundamental research by tackling challenging NLP problems and publish our findings at international conferences.
Finally, we are also involved in various stages of product development for several projects.

▶ What is the current development for the NLP at Lingua Custodia?:

Currently, our main focus is on improving our existing machine translation engines by incorporating features such as glossary and context-aware machine translation and transitioning to large multilingual models. In parallel, we are spending a lot of time on experimenting with large language models like GPT to evaluate their effectiveness in financial document understanding.

▶ What are your forecasts on the future of NLP at Lingua Custodia?:

The future of NLP at Lingua Custodia will be about expanding our NLP services by having new products that solves various problems related to finance. Precisely, we plan to focus on large language models by training and fine-tuning them on the vast amount of financial data at Lingua Custodia to make them better at understanding financial documents.