Enterprise meets language technologies

We are specialists in developing and tuning large language models based on deep learning for the private and public sector using state-of-the-art deep learning techniques.

What is NLP?

Natural Language Processing is the area of study focused on how computers understand human language, interpret it and process it to give an appropriate response.

Our mission

At LHF Labs, our mission is to make natural language processing technologies accessible to all. To achieve this, we create linguistic resources, optimise language models to reduce computational and energy costs, share our knowledge and provide advice to companies and administrations.

How can NLP help my business?

We help companies, public organisations and research centres to apply large language models (LLMs) to solve specific natural language processing problems. We can tackle a wide range of problems, ranging from machine translation, named entity detection and dialogue systems, to image recognition and other artificial intelligence applications in specific domains, such as finance, law, health and tourism. We can also help you analyse customers’ feelings and emotions regarding your products or services, allowing you to adjust your approach accordingly, or help personalise customer experiences by analysing their preferences and purchasing behaviour.

Services provided

Computation of high-dimensional language models in specific domains and languages.

Fine-tuning of language models. Distillation of existing large-scale models in specific contexts.

Application of continuous pre-training techniques for model adaptations.

Online service of language models through APIs.

Support for the exploitation of models in proprietary customer environments.

Integral consultancy for HPC, Big Data and Data Science projects, specialising in NLP.

Development of advanced conversational systems based on large language models.

The compilation, normalisation and debugging of resources for training language models.

 

Project management of resource mining and annotation for training large-scale models.

Evaluation of large-scale models.

Our process

Consulting  

We understand your needs and design an NLP business plan.

Data

We collect, clean and deduplicate corpus to suit your needs.

NLP Magic    

We adjust and prepare the models in advance so that they behave according to your specific needs. We provide constant feedback and offer recommendations to make the most of it.

Clients

The team

LHF Labs was founded in 2022 and is composed of professionals specialised in artificial intelligence, language technologies and computing in high-performance computing environments, most of them coming from the Centro Nacional de Supercomputación-Barcelona Supercomputing Center (CNS-BSC). We have experience in high-performance computing, and specialise in natural language processing and deep learning techniques in artificial intelligence. We have participated in the National Plan for Language Technologies and have published multilingual corpora with special attention to Spanish (escorpius and esCorpius-m) as well as language models (MarIA) open to the community.

Joaquin Darío Silveira Ocampo

Data Science Manager

Juan Jiménez Morillas

Business Development Manager

Antonio Miranda Escalada

Senior Data Scientist

Ksenia Kharitonova

Senior Data Scientist

Diego Saby

Senior Data Scientist

Eneko Valero Cuenca

Senior Data Scientist

Jordi Armengol Estapé

Scientific Advisor

Asier Gutiérrez Fandiño

CEO

Want to know more about LHF Labs?

Why LHF Labs?

Because we understand your business needs and how natural language processing can meet them. Because this technology is an opportunity and we can help you take advantage of it.

 

Our company offers natural language processing solutions for artificial intelligence applications in different domains (from the financial to the legal sector, from healthcare to tourism), with a special focus on machine translation problems, named entity identification, dialogue systems, question-answering systems, automatic summarisation and image recognition. We also specialise in downloading, cleaning and the preparation of linguistic and multimedia resources for the elaboration of large corpora, a fundamental step in order to obtain high quality, case-specific models. We also test them in standard and specific evaluation tasks to ensure the absence of detrimental biases.

Our solutions adapt to all types of clients, depending on the need: SMEs, large companies, public administrations and the research sector.

What can we do?

Large language models can process vast amounts of information and create contextualised responses from it. This makes it possible to create applications that understand natural language (the language people speak) and respond accurately, facilitating interaction with users.

Moreover, these systems can generate content automatically, so that complex tasks, such as translation or information search, can be performed more efficiently than with current technologies. Their application to conversational agents facilitates a better understanding of the intention and context behind a conversation, allowing users to maintain a natural dialogue and obtain the information they are looking for or trigger the desired response in a simple and accessible way for all types of people.

What is our goal?

Our goal is to help deploy these technologies and make them accessible to companies, research organisations and public administrations, providing them with powerful tools that make them more competitive and improve the quality of their services, true to our values of respect for people and diversity.

We are positioning ourselves as a catalyst for activities associated with the effective exploitation of large language models, becoming a driving force in the activities surrounding the development of artificial intelligence (AI) and, in particular, natural language processing (NLP). All this is thanks to an approach based on generating new resources for large organisations and acting as facilitators of access to them for other entities interested in incorporating these technologies into their activity.

We are involved in developing technologies, dynamisers of the sector and part of an auxiliary industry that provides services to other companies. We are a bridge between the productive fabric and language technologies. We are LHF Labs.

 

Shall we talk?

Ready to start your NLP journey?