Research Scientist (Nlp - Llm Optimization)

At LG, we make products and services that make lives better, easier and happier through increased functionality and fun. Put simply, we offer the latest innovations to make “Life Good” – from home appliances, consumer electronics, vehicle components and mobile communications to business innovations in digital signage, air conditioning, solar and LED lighting. As a global leader, we strive for greatness in product leadership, market leadership and people leadership to realize our growth strategies.

Talk about a mantra. Life’s Good with LG!

AI Research Scientist (LLM Architectures & Optimization), Conversational AI – Toronto LG is looking to push the boundaries of AI to build innovative product and service solutions to realize LG’s global vision for a connected world. As a part of this initiative, LG has set up an Advanced AI organization in North America and we are looking for an AI Research Scientist, Conversational AI for our Toronto AI Lab (located in downtown Toronto).

You will leverage your NLP domain expertise to conduct novel research in NLP and ML which will have an impact and improve LG’s existing products and technologies. You will collaborate with researchers, engineers, and product managers to design and launch AI products and solutions that help predict, personalize and transform lifestyles of LG’s global footprint of devices and users.

*1-year contract anticipated to start in September with intention to convert to full-time status

Principal Duties and Responsibilities

• Implement and improve State-of-the-art NLP models (includes novel architectural improvements)
• Participate in submitting publications to peer reviewed conferences
• Stay updated on developments in Deep Learning and NLP technologies
• Optimize models on hardware to achieve the best performance given various real time latency and power constraints
• Optimize LLM - ensuring they are performant, scalable, and cost-efficient
• Prototype and build proof-of-concepts (POCs)

Requirements

• Strong experience with deep learning frameworks (PyTorch, TensorFlow, etc.), NLP frameworks (spaCy, NLTK, gensim, etc.), and particularly Transformer-based models (BERT, ALBERT, T5, GPT, REALM etc.)
• Exposure to AWS or other cloud infrastructure
• Master’s Degree (or above) in Computer Science, Electrical Engineering, Statistics or related quantitative discipline with a focus on machine learning or related areas
• Knowledge and hands-on experience in Question Answering, Open/closed-domain Dialogue Systems, Machine Translation, Information Retrieval, Text Classification
• Ability to quickly implement models from latest research papers in NLP and ML
• Knowledge and hands-on experience in Language Modelling, Language Understanding, Language Generation, Information Retrieval, Question-answering, Chatbots, Task Oriented Dialog systems, Machine Translation etc.
• Experience working with deep NLP models and tuning their performance
• 1+ years of industry experience with PhD; OR 3+ years of industry experience with Master’s
• Working knowledge of Docker and Git

Nice to Have

• NLP related publications at top-ranked venues: ACL (or NAACL), EMNLP, SIGIR and other similar conferences.
• Experience with production environments and distributed frameworks: K8S, MLflow, Hadoop, Spark, etc.)
• Deep knowledge of modern C++ with emphasis on code generation and low level compute optimizations
• Participation in NLP-related competitions, and/or open-source NLP projects.