<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/415ba0a6-df11-4dbd-8ed0-aff811a07aeb/Logo_Submark_-_Orange.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/415ba0a6-df11-4dbd-8ed0-aff811a07aeb/Logo_Submark_-_Orange.svg" width="40px" /> The ML team focuses on two main goals: Improving the watchability of our videos and expanding the languages we can serve.

</aside>

Solving highly expressive synthetic speech - For Papercup, it’s all about serving AI dubbing, but within the ML team we’re equally obsessed with producing highly expressive speech across multiple languages. The team is research/science-oriented but all of us care about ML research that is going to have a direct impact on our product. We focus on some of the most exciting and challenging problems in speech research such as - cross-lingual prosody transfer, multilingual voice cloning, and large-scale speech model training - we’re on them all!

Just think of the last video you watched, and you’ll realise why making any videos watchable is challenging: think about the variety of content types, emotions, and characters in the videos we all know and love! And now, consider doing that not just for one pair of languages but all the pairs you can think of, and you’ll get a sense of the scale of our mission: Making the world's videos watchable in any language.

We are responsible for a research thread from conception to deployment. Depending on which part of the life cycle you find yourself in, your work could vary from data sourcing, model building, evaluation and validation to deployment. Life in the ML team is rarely dull!

Research Pillars

Modelling:

In this thread, the team balances between leveraging the latest and greatest innovations from the open-source community or academic literature to quickly iterate through ideas to develop cutting-edge research. We’ve been doing generative modelling for TTS, Voice Conversion, Lip Syncing and more since our inception.

Data:

Models are only as good as the data you feed into them. We work on a wide range of data collection efforts, spanning experiments around data scaling, data quality, and bespoke data sets to answer specific research questions.

Compute:

Big models and big data require a lot of compute. We have a central tenet in the ML team that we will never be held back by a lack of compute power. We are proud to say that we have stayed true to this tenet!

Interview Process

We want every candidate to feel prepared and know exactly what to expect. Our team always adds to these resources, so check with your recruiter if you don’t see what you’re looking for. Our Recruitment team will provide feedback throughout the process and offer you a briefing call before each interview.

The interview processes, depending on the role, can change. Please use the below as guidance. Your recruiter will communicate the full process to you when interviewing.

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/da86fd4d-7340-4866-b333-6a7f8787a966/Logo_Submark_-_Orange.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/da86fd4d-7340-4866-b333-6a7f8787a966/Logo_Submark_-_Orange.svg" width="40px" /> **Recruiter Screen

This is a non-technical meeting.**

Our recruitment team will contact you and schedule a meeting with you.

This is where you will get to know more about Papercup our culture, and the role. You will be asked questions about your skills and motivations.

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/6bb74686-11eb-47a4-bd54-896623db23c6/Logo_Submark_-_Orange.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/6bb74686-11eb-47a4-bd54-896623db23c6/Logo_Submark_-_Orange.svg" width="40px" /> Hiring Manager Screen

You will meet with the Head of Machine Learning or the hiring manager and discuss your experiences with research and building ML products. This meeting is a chance for both sides to determine if there is a potential mutual fit.

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/2b3bcb7d-a37b-4a04-875b-71cc6538d190/Logo_Submark_-_Orange.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/2b3bcb7d-a37b-4a04-875b-71cc6538d190/Logo_Submark_-_Orange.svg" width="40px" /> Role-Related Technical Knowledge Interview

There are two parts to this interview:

Coding Interview: The team will gauge if you are comfortable working with PyTorch frameworks or similar frameworks.

Paper Interview: This will be a 1-hour interview. We will send you a choice of papers and ask you to present one back to us. As a team we frequently read papers in the speech literature you will be asked to present to the team and spark a discussion on how to think about the piece of work in the context of our use case. We’re looking to assess your ability to read, present, and critically think about the kind of research that we do. Hopefully, we can get you excited about our challenges as a research team.

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/7aeee626-eecf-43ed-a5ec-15761eea2df2/Logo_Submark_-_Orange.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/7aeee626-eecf-43ed-a5ec-15761eea2df2/Logo_Submark_-_Orange.svg" width="40px" /> **Values Interview

This is a non-technical interview.**

This is a behavioural-based Interview where we will ask you questions based on real-life experiences that align with our values.

Our values guide our behaviour and help explain our actions.

The question you will be asked will relate to a specific situation or challenge you may have had.

</aside>

<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/f3017f4e-9b1a-47b8-a8fb-54bee477b04a/Logo_Submark_-_Orange.svg" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/241ec0d1-0af5-48fd-815f-bf3c34a0b982/f3017f4e-9b1a-47b8-a8fb-54bee477b04a/Logo_Submark_-_Orange.svg" width="40px" /> **Leadership Interview

This is a non-technical Interview.**

We want to evaluate how we can guide and support you in the position you are interviewing for and your future progression at Papercup.

We will dive deeper into your long-term motivations, goals, and passion for what you do.

</aside>