Senior ML Engineer, Personalized Noise Cancellation
🦉

Senior ML Engineer, Personalized Noise Cancellation

About Krisp:

At Krisp, we are developing and launching applications that enhance professionalism and productivity of everyone. Krisp mutes background noise bi-directionally during calls and is the only software available in the market to do so as an application, becoming a virtual microphone and speaker within Windows, Mac OS, and VDI. We are out front leading this market, with partners and customers including Sitel, Concentrix, Discord and many others. Our team is defining the future of how remote employees vastly improve their voice and video communications.

Krisp is an award winning product, loved throughout the world. In 2020, Krisp has been recognized as:

  • Forbes AI 50 most promising AI startup
  • Forbes Cloud 100 Rising Star
  • TIME 100 Best Inventions
  • Gartner Cool Vendor

Location:

Armenia, Remote

We are a remote-first company, so you can work from anywhere. However, based on the business need some roles are considered for specific locations.

The Role:

We are looking for a talented Senior ML Engineer to join our experienced research team at Krisp. You will join a team working on a unique technology around the Personalized Noise Cancellation problem. In this problem one tries to filter the voice of a preselected person from a mixture of various voices and noises, with low latency. This is similar to the famous cocktail party effect, with the difference that one should obtain the voice of the target speaker in its entirety and with low latency.

As a Senior ML Engineer, you will support the complete life cycle of a research product. This includes software development, gaining domain-specific knowledge related to the problem, research into state-of-the-art approaches to similar problems, data preparation, and training of cutting-edge ML/DL algorithms, determination of testing priorities, fine-tuning of models and other functions.

What You'll Do:

  • Product Development:
    • Be result oriented and develop an excellent understanding of the product from the standpoint of the end user. Form a holistic view of the product in development. Contemplate about the tradeoffs between technical, analytical and product needs and lead to solutions that take all such requirements into consideration
    • Research to find or create appropriate datasets. Carry out organization, preprocessing, calculating statistics and if necessary changes in representation
    • Considering the product use cases, generate the corresponding data to train deep learning models starting from raw data. Develop and apply various domain specific augmentations when necessary
    • Research, implement in detail and train appropriate ML/DL algorithms and tools for the Personalized Noise Cancellation problem
    • Work with QA to conduct insightful and actionable tests and develop automated tests whenever possible
    • Based on QA test results and user feedback, reconsider the datasets, fine tune the data generation to use cases, retrain, test again and iterate
  • Professional Software Development:
    • OOP methodologies, attention to interfaces, key design patterns
    • Version control
    • Using right data structures and algorithms to achieve low time and space complexity 
    • Vectorized algorithms and calculations, allowing good usage of the CPU cache, optimal memory usage and other resources, multiprocessing and multithreading
  • Gaining Domain-Specific Knowledge:
    • Studying the digital representation of audio and video data
    • Digital Signal Processing
    • Human Speech Structure
    • Relevant manipulations of speech and video data
  • Simultaneously being well informed about the top of industry trends and being innovative in research by proposing new approaches based on convincing scientific and technical arguments.

What We’re Looking For:

  • MSc or PhD in Mathematics, Physics or Computer Science
  • Proven experience as a Machine Learning Engineer or a similar role
  • Knowledge of mathematics, particularly linear algebra, probability, statistics and optimization
  • Understanding of training data structures, data modeling and software architecture
  • In-depth understanding of mathematical ideas underlying ML algorithms
  • Good knowledge of data structures and algorithms
  • Ability to write production-level code in Python
  • Familiarity with machine learning frameworks (e.g. TensorFlow, Keras, PyTorch) and relevant libraries
  • Familiarity with voice and digital signal processing is a plus
  • Analytical and problem-solving skills
  • Excellent communication skills and team spirit.

Benefits at Krisp:

  • Stock options
  • Learning and development assistance
  • Health, dental and vision insurance
  • Paid Time Off and Sick Time Off
  • Paid Parental Leave
  • Maternity Leave Top-up
  • Gym reimbursement
  • Referral bonus program
  • Work from forest (remote work initiatives)

Krisp is an Equal Opportunity Employer:

All applicants are considered without regards to race, color, religion, national origin, age, sex, marital status, ancestry, physical or mental disability, veteran status, gender identity, or sexual orientation. We do not tolerate discrimination or harassment of any kind. All employees and contractors of Krisp treat each other with respect and empathy.

How to Apply:

All interested candidates are encouraged to submit their applications.