As a Research Scientist at Cobalt Speech and Language, you will have the opportunity to architect and implement novel natural language processing (NLP) products as Principal Investigator on U.S.-government-funded projects. Cobalt is a collaborative team of world-class speech scientists and engineers working from their home offices around the world. You will engage with our team to leverage our state of the art speech recognition and dialogue system capabilities. We are seeking a researcher who can synthesize the latest NLP techniques in practical real-world applications, shepherding projects from conception and design, through building and testing, and into deployment.

The successful candidate will be:

• Familiar with machine learning (ML) techniques, ideally in NLP applications. Experience in at least one of the deep learning toolkits (TensorFlow, PyTorch, kaldi, etc.) is a plus.
• Experienced in selecting the appropriate ML algorithm for a task, understanding feature extraction, neural network architectures, and appropriate cost functions.
• Able to perform data and error analysis in order to improve ML models and understand their shortcomings.
• Experienced with real-world data sets and the preparation of such data for ML training.
• Proficient at coding in one or more scripting languages (e.g. Python, bash scripting).
• Willing to learn to code effectively in Go. Prior Go experience is not required, but adaptability and demonstrated ability to learn a new coding language is essential.
• Self-motivated and proactive.
• Able to collaborate effectively with a virtual team.
• Authorized to work in the U.S. (citizen or green card holder)

Education: M.S. or B.S and equivalent experience in a relevant field (Computer Science, Computational Linguistics, etc.)

Are you looking for a high-impact position at an innovative company doing cutting-edge research and development? Do you thrive on creative problem solving and continuous learning opportunities? Do you enjoy working with a close-knit team in a quirky, nerdy, supportive company culture?

As a Software Engineer at Cobalt Speech and Language, you will have the opportunity to collaborate with world-class speech scientists to develop specialized software allowing our customers to solve novel problems using voice and natural language technology.

The successful candidate will be:

• Excited by the prospect of learning cutting-edge technologies in speech recognition, natural language processing, and machine learning
• Proficient coding Go, Python and/or C/C++
• Experienced in developing enterprise software products
• Versatile and interested in the full software stack, from APIs to algorithms to build and deployment scripts
• Experience with knowledge graph databases is a plus
• Customer-focused
• Self-motivated and proactive
• Able to collaborate effectively with a virtual team
• Able to communicate well in English, both spoken and written

Job responsibilities:

The primary responsibility for this position will be to extend and improve Diatheke, our dialog product for building voice user interfaces (VUIs) like smart assistants, and use Diatheke to solve challenging problems for specific customer projects. This product brings together nearly the whole suite of Cobalt capabilities, from ASR to NLU to TTS, so this is an important, collaborative role. 

Development is primarily in Go, using gRPC and protobufs to communicate with microservices written in Go and C++.  Prior experience with Go is not required as long as you can demonstrate proficiency in learning a new language.

Experience with VUIs is a plus, but also not required. What’s required is eagerness to learn and push the edges of what’s possible in this rapidly developing area.