CarLabs specializes in building conversational experiences in the automotive vertical. We're building out a platform for Messenger bots, Alexa skills, Google Assistant/Home, and web-widgets. Our goal is to change the way people find and learn about the best car for them. We want to allow our clients to create unique and creative experiences with responses backed by automotive data.

At CarLabs, you will work on a small, but dedicated, team to solve challenging problems in Machine Learning - primarily Natural Language Processing. We make software that can market and sell a car (on various messaging and voice platforms) as well as answer owners' questions and respond to their concerns - and though we are the best at it, it is by no means a solved problem. That means we are constantly evaluating new techniques, because no one knows what the best possible performance is. Should we try to add a Q&A model to the ensemble, or rearchitect the classifier for higher accuracy? Recently, the paper Generating Wikipedia by Summarizing Long Sequences got some great results using "extractive summarization to coarsely identify salient information and a neural abstractive model to generate the article" - should we explore this, or is our time better spent mining the logs to see if we can identify new classifications for customer input?

As a Lead, you will make these calls. You will help define the ML roadmap and then be the owner of modeling and delivery of all Machine Learning products. You will become familiar with the latest research on NLP, and implement solutions that range from experimental to tried-and-true. A typical week would entail brainstorming with product on features, prototyping models and improving model features with a small team, and collaborating with engineers to deploy ML pipelines and products.

We are strong believers in choosing the right tool for the job, and nothing we have done so far is set in stone - you will have a very high degree of autonomy. You won't be on your own, though - our engineering team and data science team are very close, and you will have their support when needed to build tooling. In addition to their support, you will have the engineering team's attention - there is lots of demand among the team to learn about your field.

• At least four years of quantitative experience with large data-sets, from prototyping to business impact
• Deep understanding of machine learning and statistical methods with their underlying theory and math
• Knowledge of, or the desire and ability to quickly become familiar with, how RNNs are used in Natural Language Processing
• Experience building, deploying and demonstrating business value from predictive models and data products
• Strong Python skills
• Experience with some of the following: PyTorch, Keras, TensorFlow, MXNet, Caffe, Theano, scikit-learn, gensim, Stanford CoreNLP, Spark
• While we assume you have a Master's or Doctorate in a quantitative field, for a candidate with the right experience and background, we can make an exception
• Excellent communication and problem-solving skills. You will frequently interact with the executive team, some of whom do not have a technical background. The ability to discuss models, assumptions, and outcomes in business terms is indispensable. This includes the ability to use data visualization tools (which ones is up to you)

All your information will be kept confidential according to EEO guidelines.

Unfortunately, we are not able to sponsor visas at this time.