San Francisco, CA

  • Coding principles
    Whiteboard before first line of code, Test-driven development, Bleeding-edge stack, Pair programming, Everybody can push to prod
  • Work/Life Rhythm
    Remote as needed, No taking work home
  • Modern Tech Stack
    GitHub, Travis CI, Pandas, TensorFlow, React, AWS, Python

Our Mission

At Huma, we help our customers get previously unobtainable, instant insights across their sea of documents and data by simply asking questions and getting answers. As a startup, we already have “delighted” customers because the insights the platform is generating can be used to bring life-saving drugs to patients faster and with greater accuracy. Imagine conversing with your data and instantly getting straightforward answers! Add magical usability: Go from existing tap, click, and swipes to a “Just Do What I Want” user experience. Connect Disparate Systems: Huma automatically connects different data silos through human language. Self-Service Infomatics: Get instant answers from your data.

Tech Stack

Python, Pandas, PyTorch MySQL SpaCy AWS S3, Dynamo DB, RDS, Redshift, Elastic Search We are building novel solutions to very difficult problems, such as : entity/relation extraction, normalization, semantic search, ranking, oncology-aware IR, text summarization. Our tech stack involves modern deep learning approaches to NLP: data augmentation, text classification, ambiguity disambiguation, representation learning, etc. We love to keep up with recent advances in deep learning NLP, NLU and ability to modify standard algorithms and implement in production quality code. It's really important to us that every engineer on our team has real world experience with developing and training neural network models. We care about writing clean and efficient code, as well as practical experiences in building production quality application.

What Our Team is Reading

Xz format inadequate for long-term archiving (2016)

Team Q&A

  • How do you apply machine learning?

    Greg Kostello

Core Values

We value solving problems that have a real impact. On our team, you will leverage your skills to develop new insights into unstructured data using the standard as well as very new, advanced NLU (natural language understanding) frameworks. You are comfortable both defining problems and solutions and transforming data through a data pipeline into forms that can be readily analyzed. You know that data is “messy” and that a multiphase approach to wrangling the answers from data is part of the job. The principal data scientist may also have scientific expertise in the field of study, often to great advantage for the purpose of this role, though not necessarily at the level of the experimental scientist/physician/customer. Alternatively, the principal data scientist may have little or no scientific expertise in that particular domain but bring an extremely advanced set of skills in the areas of machine learning, deep learning, statistical modeling, data engineering, software engineering etc., such that they are generally advantageous members of a project or subproject team regardless of the scientific domain of the problem.

Vacation & Benefits

We have a flexible, unlimited vacation policy. We trust everyone to make the best decisions about when to take off and recharge and to get a ‘thumbs up’ from their teammates to make sure projects continue smoothly.


  • Psychological safety
  • Intellectually curious
  • Always be learning
  • Freedom & autonomy
  • Trust our intuition
  • Empathy with customers

Cool Tech

  • Artificial intelligence

Day in the Life

  • Weekly executive Q&A
  • Globally distributed dev team
  • Daily lunch together
  • Continous feedback
  • < 5 hours of meetings
  • Private offices
  • Great for parents
  • Do Not Disturb time blocks

Team members

  • 12 core team members

Greg Kostello


  • Mac
  • Sublime
  • Logitech
  • Python, Javascript
  • Settlers of Catan
  • Civilization

Machine Learning Data Scientist

San Francisco, CA・Mid-Level・Full Time


More Companies You May Like

2 jobs


We help our customers make the most of what they have. Our technology connects people, devices and insights to better manage energy and water resources, all around the world....

Liberty Lake, WA & Globally distributed team
2 jobs

Scientific Games

Our mission is to empower our customers by creating the world’s best gaming and lottery experiences....

Globally distributed team
3 jobs


We develop open-source software for a world dependent on Big Data. With Cloudera, businesses and other organizations can now interact with the world's largest data sets at the speed of thought — an...

Palo Alto, CA