Genei build AI tools that learn from human feedback.

NLP Research Engineer

$100k - $120k / 0.50% - 1.50%
London, UK / Remote
Job Type
3+ years
Connect directly with founders of the best YC-funded startups.
Apply to role ›
Jack Bowen
Jack Bowen

About the role

About us

At Genei, we build AI systems that learn from human feedback. We’re a highly motivated and talented team of scientists and engineers building at the intersection of massive scale language modelling, reinforcement learning and AI alignment.

Our consumer product,, delivers state of the art search and summarisation natural language processing to over 4,000 students and researchers worldwide. As the users interact with the app, our online learning algorithms adapt to provide the user with the optimal experience. The algorithms we’ve developed have massive potential outside of our consumer app, and we’re starting to offer them as a service to businesses, allowing them to leverage their users to improve their AI capabilities over time.

The role

Whilst the role is quite open ended and requires high levels of self motivation, a typical workflow might be:

  1. Read and understand the latest literature
  2. Use this understanding to come up with novel ways to improve Genei’s ML
  3. Design and run live experiments to test hypotheses
  4. Transform the findings of these into production code and features

What we’re looking for

The ideal candidate is someone who can demonstrate:

  • A genuine passion for Genei’s research
  • The ability to set goals and work independently towards them

We’d expect the following experience:

  • 3-5 years experience working with large scale deep learning models and transformers

Strong candidates would have experience and interest in at least 1 of the following areas

  • Reinforcement Learning applied to NLP
  • Large distributed language model training
  • Multimodal transformers on documents

About the interview

Interviewing for a role at Genei

  • Genei is looking for engineers who exhibit a high degree of ownership and autonomy; enjoy a steep learning curve and are able to ship and iterate quickly.
  • Our interview process aims to find engineers of this description who will thrive and enjoy working at Genei.

Outline of interview process (~3-4 weeks from first contact)

  1. Informal Phone Call
  2. Technical Screen
  3. Interview
  • Team Fit (non-coding)
  • Debugging Task (Coding)
  • Whiteboard Task (Coding)
  • Experimental Design Task (Coding)
  1. Feedback

Informal Phone Call

  • This is a chance for you to find out about the role and evaluate whether it's for you.
  • We are looking for people with a strong interest in creating high-performance implementations deep learning algorithms.

1. What is it like to work at Genei?

  • We are a small and highly collaborative company.
  • We face new challenges frequently and adapt quickly to learn on the job and solve them.
  • We are hard working; evidence driven and thoroughly enjoy coming to work on a Monday morning.

2. What has your past experience been?

  • Do you have relevant experience working with any of the following:
    • Large distributed language model training
    • Reinforcement learning applied to NLP
    • Multimodal transformers applied to documents
  • Are you looking for a paid opportunity to level up your skills in the space?
  • Are you looking for a strong sense of ownership and autonomy?
  • Have you worked in a small fast moving team within a company or startup?

Technical Screen

  • This will be a first pass to assess basic skills and determine whether to progress to a more in depth interview.

The Interview

  • The interview consists of 5 parts.
  • We may choose to short circuit decision making to hire / not hire a given candidate after any of the stages below.

Team Fit (Non-coding)

  • This is a personality interview and will aim to understand more about your past experiences and how you work in small teams.

Debugging Task (Coding Task)

  • All engineers at Genei will be working at some point with existing codebases.
  • In this task we will give you some machine learning code to debug and fix.
  • We are interested in hearing you talk through your process and thinking as you do this.

Whiteboard Task (Coding Task)

  • This will involve working on and talking us through the solution to a machine learning problem followed by a discussion around the pros and cons of your implemented solution.

Experimental Design Task (Non-coding)

  • We are looking to understand your ability to design, build and test the performance of new techniques and models.
  • During this we will present you with an open ended experimental design question and expect you to propose and talk us through a possible approach.


  • We will aim to get back to you in 3-4 weeks with a decision and feedback
  • The exact time to complete the process will depend on availability of the founders

Why you should join Genei

At Genei we build Ai tools that learn from human feedback. We offer this as a tool to consumers and a service to businesses. Our mission is to bring aligned, accountable and personalised AI to every knowledge worker on the planet.

We are a small, hard working evidence driven team. We love new challenges and adapting to overcome them. While working here we we expect and encourage everyone to develop a deep curiosity, passion and pride for their specific craft.

Team Size:5
Location:London, United Kingdom
Jack Bowen
Jack Bowen
Thomas Foster
Thomas Foster
Adrien Wald
Adrien Wald