Verbit

Description

We are looking for a versatile and motivated Speech Researcher/Engineer to join our Research team. This role is centered around acoustic modeling, including ASR, speaker modeling, and acoustic event detection, but extends to a broad range of machine learning and AI tasks.

You will work on high-impact projects involving the training and optimization of deep learning models, conducting evaluations, improving production systems, and contributing to research-driven initiatives. A key part of this role is the ability to work with complex, non-trivial codebases, especially in areas like speaker identification and diarization.

As a Speech Researcher/Engineer, you will:

• Develop and improve ASR, speaker modeling, and acoustic event detection systems.
• Analyze and optimize performance of existing speech models (WER, diarization error, latency, etc.).
• Dive into and extend complex, production-grade codebases for training, inference, and evaluation.
• Contribute to dataset preparation and model training pipelines.
• Participate in broader ML/AI efforts as needed, beyond speech-focused tasks.
• Collaborate with cross-functional teams across research, engineering, and product.

If you want to join our journey, you’ll need:

• B.Sc. or M.Sc. in Computer Science, Electrical Engineering, or a related field.
• Solid background in machine learning, deep learning, and speech/audio modeling - 3+ years of experience, preferably 5+.
• Strong Python skills and experience working in Linux environments.
• Familiarity with deep learning libraries (e.g., PyTorch, TensorFlow) and training workflows.

It will be even better if you have:

• Experience with speaker recognition, speaker diarization, acoustic event detection.
• Familiarity with cloud infrastructure (e.g., AWS, OCI).
• Exposure to MLOps

What makes Verbit unique?

Verbit’s global team is united in its mission: to make all verbal information and experience accessible, insightful, and useful.

Powered by our award-winning AI technology, Verbit helps businesses, organizations, and individuals of all sizes make words work—whether it’s a legal deposition, a content creator’s latest campaign, or a major global event.

With a global network of human experts and a continually evolving proprietary AI engine, Verbit ensures exceptional results while scaling to meet any need.

We're building a world in which all speech can be seamlessly converted into meaningful actions. Join us from our offices across the United States, Canada, Israel, and Europe.

Do you have Verbit DNA?

Verbit’s people are committed to “winning together” through constant collaboration to have an impact on the world. They share a “do good” mentality and apply it daily in their work.

We’re a group of:

• Tech-savvy individuals who are always open to growth and learning
• Adaptable and flexible people who thrive in a fast-paced environment
• Creative minds who rethink and question how to outperform past results
• Effective communicators who can promote and represent Verbit’s tech and brand