Senior ML Scientist – Speech Processing at AVA

About the job

Reinvent speech recognition technologies to empower us collectively

AI is an incredible thing, especially when we put it to the service of humanity — and not the other way around. We love hearing incredible applications of computer vision, from detecting cancer by looking at your eye or saving your life by making driving autonomous & more secure. But speech recognition applications? They’ve stayed pretty much at the gadget level, inundating us with talking alarms and smart home assistants.

Ava is the exception to the rule – the ampersand-shaped peg in a rather boringly square world. We apply modern speech processing and speaker identification (e.g., diarization) technologies so we can help 450M deaf & hard-of-hearing people live a fully accessible life. With Ava, they can understand what everyone else is saying in any group conversation, anytime. With five asynchronous audio streams arriving at your doorstep, how do you even decide what audio should go into your ASR? How to process them collaboratively to improve the ASR transcription quality?

A difficult challenge to solve? Yes, and that’s what we love. What really keeps us up at night though, is the thrill to be impacting day to day people in a meaningful way. And see the smile on our users’ faces when they use our product.

Don’t waste any more of your precious gray matter in politics-laden research teams in big companies, and come instead shine your light where the world can see the impact of what you make!

What will you do as a Senior Machine Learning Scientist at Ava?

  • You’ll own one of the Ava’s AI team research objectives to develop solutions for real-world applications of our core technology.
  • You’ll keep an eye out on the state-of-the-art research, through publishing your work at the high-level conferences or networking with peers, and think of new contributions to improve the state of our stack.
  • You’ll supervise master students and possibly PhD student. You’ll advise other AI team members and will collaborate with them.
  • You’ll help improve the performance and reliability of our cutting-edge AI systems — such as improving the speed and responsiveness of our deep neural networks.
  • You’ll work with the rest of the AI team to build and maintain robust internal machine-learning tools and processes, with a focus on our Machine Learning models. Think of projects such as detecting new voices in a 0-shot learning setting and separating overlapped speech.

You’d be perfect for this role if:

  • You aspire to be a pioneer in the field, and do what is necessary to make things work in real world situations.
  • You hold a PhD in machine learning esp. deep learning or in signal processing, or in a related field.
  • You have 5+ years of research and development experience (preferably in industry) after completing your PhD.
  • You have a strong publication record with some contributions remarked by the research community.
  • You have experience in online/real-time audio processing, multichannel audio analysis/processing, speaker diarization, speaker identification, denosing, dereverberation or source separation, with the methods and the representations involved.
  • You are fluent in python and have had exposure to tensorflow and/or pytorch.
  • You’re of the persistent, yet open-minded and collaborative type: you can own and pursue a research agenda and autonomously carry it out, but appreciate working in a team.
  • You are able defining relevant research topics/directions and supervise/guide students or other team members.
  • You care about the strategic and business implications of anything you build. You’re not just going after cool stuff — you understand the balance between craft, speed, and the bottom line.
  • You are excited when the algorithms you have designed and developed are running in the product and are effectively used by the clients.
  • Bonus: You have past experience creating high-performance implementations of deep learning algorithms.

What’s special about this role?

  • Actually change lives at an unprecedented scale: How often do you hear about apps that make people cry of joy? It is really for us a unique opportunity and privilege to be able to meaningfully improve the lives of 100,000s of people – and yet to still be at the very beginning of our mission!
  • Work with some of the best people in the world: We have an incredibly talented and passionate team that is a lot of fun to work with. We’re still super small and have accomplished some things that were thought impossible!
  • Tackle our most interesting and impactful problems: Our AI team is still small, and people wear many hats. You’d jump between machine-learning models, infrastructure, internal tools, process — participating in every phase from inception to implementation. Absolutely no boredom.
  • Join us at an incredible time: We’re well-funded and hit product-market fit, which gives us a huge green field to work with. You’d join at the perfect time to shape what we build and how we grow, so we can create a more inclusive world.

Location & Remote work

  • The AI team is based in Paris. We’re going for an intentional use of in-person time. If you are based in/around Paris, you can expect to spend more time in the office than if you are further away. But in any case, this position will require a certain amount of in-person time for collaboration. Let us know what your ideal arrangement would look like!



Job Category: Research Scientist
Job Type: Full Time
Job Location: Europe