AI Scientist - NLP - Relationship Extraction

What's the role and how do you fit in?

BenevolentAI harnesses artificial intelligence to accelerate scientific discovery by making sense of highly fragmented information to develop new medicines for hard to treat diseases, using AI as a force for good. Valued at $2bn in 2018, we are the largest independent AI company in the world.

We are seeking an AI Scientist / Machine Learning Researcher with experience in empirical research in natural language processing (NLP) and deep learning to join our London office. You will be responsible for developing and executing a research agenda around the core problems we work on at BenevolentAI - scientific discovery from structured and unstructured data sources.

What will you be accountable for?

Examples of NLP challenges we are solving on daily basis:

  • Extracting complex facts: Simple binary relationships are not a very good representation of biological facts, which often have a complex, nested structure. This is a very challenging problem for current state-of-the-art systems.

  • Modelling the evolution of novel scientific knowledge: We are very interested in modelling the evolution of a scientific fact within a research community, from initial hypothesis to well-established fact. The idea of ‘novelty’ is very important to us and has a lot of overlap with the work we are doing in knowledge-base inference.

  • Identifying the evidential basis of scientific knowledge: We would like to characterise the evidence supporting scientific statements - a fact with strong experimental support is very different from a speculative or controversial fact.

  • Finding signal in noisy, unlabelled or weakly labelled datasets: Existing annotated datasets are often limited or have been constructed for related but not identical problems. We are interested in exploiting and combining various sources of potential signal in our models.

  • Incorporating feedback from in-house drug scientists in NLP models: The output of our models is consumed by our in-house drug scientists to help them find new drugs. Our challenge is to extract the most relevant information to them. We collaborate closely with our drug scientists for the design, evaluation and iterative improvement of our models.

  • Developing scalable models: so they can be incorporated in our end-to-end pipeline which regularly processes millions of biomedical documents.

What skills, experience, and qualifications do you need?

  • An advanced degree (PhD, MSc) in computer science or a related field with a clear focus on empirical research.

  • Practical experience with ML methods for NLP (e.g. relationship extraction or named entity recognition) and relevant research for these topics.

  • Ambition to publish your research results in top tier conferences and journals

  • Knowledge of modern tools for ML such as Tensorflow, Pytorch etc.

  • Experience building products is a plus.

    If you would like to join our growing group of scientists / researchers to have direct impact in the goals of our company by advancing drug discovery, we would love to hear from you.

About BenevolentAI

BenevolentAI, founded in 2013, is an advanced technology company focused on accelerating the journey from data to medicines. It is the world’s only technology company with end-to-end capability from early discovery to late stage clinical development. The company is HQ’d in London with a research facility in Cambridge (UK) and further offices in New York and Belgium.

The ‘Benevolent Platform’ is a unique machine intelligence technology system built to mine new knowledge from vast quantities of biomedical data, propose treatments and design drugs to enable its world leading scientists to bring new treatments to patients faster.

We are working on applying tech to real problems, and see real outcomes and the fruits of our labour by working on a meaningful mission. We do our job ‘Because it matters’ and live by the philosophy that unconventional thinking together with purposeful technology can have an impact on humanity.

The working environment is agile and we work in cross functional teams. We encourage a culture of learning, developing and challenging the status quo to foster dynamic, entrepreneurial behaviours, innovation and a ‘fail fast’ mentality.

Alongside all of this we can offer excellent benefits (learn more at, a global reach and the ability to work with the best talent in the industry.

Please note that by applying for a job you agree that we will collect and process your personal data submitted in your job application in accordance with our Privacy Policy.