Reference : FRA16939

INTERNSHIP - NLP Deep Learning For Meeting Summarizer (F/H)

  • France LA DEFENSE CEDEX
  • Internship
  • Posting date : 11/14/2022

Position description

We are looking for someone that is passionate about natural language processing (NLP) and deep learning, someone that likes to push the boundaries of what is possible, and want to solve hard problems on top of unique datasets. You will be responsible for implementing, customizing, and inventing NLP deep learning techniques for some of the most challenging datasets. 

You will be part of the Saint-Gobain Data & Analytics organization which aim is to use data for providing operational excellence and impactful solutions in the context of the direction towards carbon neutrality. The main pillars of this program are data governance, modular data platform, AI@Scale and MLOps.

You will be integrated to the AI & Data Analytics team which main scope is to deliver transformational insights with value impact on all related businesses of Saint-Gobain. 

Role

You will be part of a team developing a French-English meeting summarization tool. You will use state-of-the-art NLP techniques to build robust and accurate models. Depending on the performance of the model developed, we could publish a research paper in some peer-reviewed conference to explain our approach and share the model with the NLP community (HuggingFace Hub).

Missions

  • Read, present, and implement research papers in the areas of NLP and deep learning.
  • Modify and/or invent state of the art techniques to solve real-world problems and apply them to tasks such as Natural Language Understanding, Summarization, Text Classification, Intent Classification, Question Answering, etc. to optimize performance on unique datasets.
  • Work closely with engineering teams to push innovative learning models to production.
  • Propose and implement new ideas/directions in NLP/Deep Learning that can effectively solve customer problems.

REQUIRED QUALIFICATIONS

•    Experience with Python and a Deep Learning framework (Pytorch, Tensorflow).
•    Experience with the latest NLP models such as BERT, GPT-2/3, T5, BART, etc. 
•    Experience with deep learning NLP toolkits such as Hugging Face, spacy, etc. 
•    Experience with vision-and-language multimodal pre-training and fine-tuning. 
•    Proficiency with any of the major Deep Learning frameworks and libraries. 
•    MS or equivalent in Machine Learning, Computer Science, Statistics or related fields. 
•    Publication(s) at peer-reviewed conferences or journals. 
•    Excellent communication and collaboration skills. 
•    Student in 4th or 5th year of engineering school or university, specialized in Machine Learning, Computer Science, Statistics or related fields.

Who are we ?

At Saint-Gobain, our purpose is to 'Make the world a better home.' We, as Saint-Gobain, are a top global employer, a historical company in France and Europe, and a worldwide purpose-driven leader in the building materials industry and lightweight constructions. We have a long-term ambition to make our planet a better home. We are involved in building a more sustainable and inclusive world to help protect the environment and leave the world to future generations better than we found it.

We are both manufacturers and distributors. Our products are everywhere in our living places and daily life: in buildings, transportation, infrastructure, and in many industrial applications. Together with and for our customers, we design materials and solutions that positively impact everyone's life and provide well-being, quality of life, and performance while caring for the planet.

top_employer_logo_2023
 Happy trainee 2022