Unlikely AI is an early stage Artificial Intelligence start-up based in London.
Unlikely AI was founded by William Tunstall-Pedoe, who previously served as the principal project manager for Amazon’s AI voice assistant Alexa. Tunstall-Pedoe’s earlier startup Evi Technologies was key to the natural language understanding and question-answering seen in the Amazon Echo and Echo Dot.
The application of the technology being developed by Unlikely AI is not clear at the moment. The company has, however, released a showcase of its AI capability in the form of an AI called Ross. Ross featured in an app called Crossword Genius, in which Ross solved cryptic crossword puzzles.
This is a key opportunity to join as an early employee, to make a significant impact on the fortunes of the business and help define the company, culture, product and software engineering practices.
The Role
- As a Senior Applied Scientist you will help in providing technical leadership within Applied Science and across the company.
- Convert cutting edge research into real products
- Design, build and experiment with cutting edge technologies
- Drive the design, development, and execution of scientific research projects
- Write software that is built and deployed in production systems
- Communicate complex solutions to colleagues
- Analyse & inspect large scale datasets
- 6+ years of hands-on experience in deep learning
- 2+ years of experience applying deep learning to NLP for example Large Language Models LLMs,Sequence to Sequence models such as Neuro - Machine Translation (NMT) and seq2seq parsers.
- Experience utilising & deploying transformer modelsDeep knowledge of machine learning fundamentals
- Strong coding skills in Python, including the use of pytorch or tensorflow
- Enthusiasm to learn and get up to speed with cutting edge technologies which you may not already be deeply familiar with
- Strong verbal and written communication skillsCloud
- A clear track record of mentorship and coaching others
This job is already closed and no longer accepting applicants, sorry.