We're sorry, but this job has been placed on hold. See other open jobs at Impermium
Not the right job?
Principal Data Scientist at Impermium in Redwood City, CA
If you are a Machine Learning guru or a Data Science magician with strong hands-on engineering experience, please read on!
- Locality Sensitive Hashing!
- Random Forest Ensemble Classifiers!
- Stochastic Gradient Boosting Distributed Decision Trees!
As Impermium continues to build out its Internet-scale user-generated content classification system, we're looking for someone whose passion lies in the invention and application of cutting-edge machine learning, data-mining and data analysis algorithms. Our systems already classify tens of millions of transactions every day, looking for spam, abuse, fraud, and other bad behavior -- as Principal Data Scientist, you'll join our team to help guide and shape our algorithm and classifier development.
Most large-scale abuse classification systems break down due to non-I.I.D. distributions, an over-reliance on exhaustive “ground truth” training corpora, and an adversary who continually adapts to specific weaknesses in the classifier. Your challenge is to guide pioneering work to overcome these historical limitations.
The ideal candidate is a highly knowledgeable, all-star computer scientist, with a strong background in machine learning, data mining and statistics. In order to be successful, you must have previous hands-on experience turning conversations into prototypes and prototypes into products -- ideally in a startup environment. You must also have a strong track record of creating novel and significant intellectual property for your company. Last but not least, you must be a confident public speaker, willing and able to passionately represent your and Impermium's ideas on stage.
You'll fit right in if:
- You are a self-managed, high-energy individual with a love for data
- You possess exceptional communication skills with the ability to clearly articulate your engineering and product ideas with technical and non-technical team-members, as well as customers
- You have a passion for mentorship and enjoy explaining and simplifying complex concepts to implementors on the team
- You are absolutely confident in your ability to design scalable, principled, practical classification and clustering algorithms that operate within the near-real-time constraints of the abuse and fraud domains
- Proven ability to develop and execute sophisticated data mining & modeling projects
- 5+ years experience creating prototypes that are shipped to market (production systems)
- 5+ years experience in software product development - with the core focus being on mathematical and / or statistical algorithms
Well-versed in a modern general purpose programming language such as Python
Well-versed in one or more analytical platforms or frameworks such as R
Experience working with "Big Data" systems and platforms such as Hadoop and Mahout
Well-versed with unix command line tools and one or more scripting language such as awk/perl/shell
- NLP (natural language processing) experience is a plus
- Publications and/or patents are a plus
- Ph.D in Computer Science or related fields
Impermium offers you:
- A chance to build a crucial component of Web 2.0 infrastructure: the defense against spam and abuse
- A dynamic, technology-driven work environment in our brand new office, convenient to 101 and Caltrain
- A highly influential and visible role with direct impact on foundational product and engineering direction
- The opportunity to work alongside a highly talented, experienced founding team