Overview
What we do
Here at Mavenoid, we are building a virtual product experience that helps customers configure, use, and troubleshoot their products. With some of the world’s most iconic brands as clients and $10 million in funding, we're reinventing technical support. Our goal is to help customers repair more and replace less, get more out of their products, and never wait on hold again. Founded in 2017 by a team from Palantir and Google, Mavenoid has quickly established itself as the leader in tech-support automation. We are located in Stockholm but doesn't mean you have to be - this is a remote-friendly workplace.
What you'll do
- Collaborate with stakeholders and other engineers to conceptualize, build and test new product features.
- Apply the right tools for the job and solve business problems with real-world challenges.
- Develop knowledge representations and virtual-agent behavior combining end-user, expert and machine-learning values.
- Provide model explanations and apply structure learning to refine product knowledge.
- Learn desired behavior from examples using causal reasoning or imitation learning.
Qualifications
- At least two years of industry experience in AI/ML/data-science roles
- Fluency in Python and a deep-learning framework such as TensorFlow or PyTorch.
- Accomplished in probabilistic models and natural-language processing.
- You love working on hard and valuable problems, no matter how challenging.
- You can solve a wide breadth of problems in more direct or elegant ways than most.
What we offer
The upsides of joining a small company early: outsized impact, real ownership and accountability, hard problems, faster learning, quicker decisions, less bureaucracy, and meaningful equity combined with fair comp.
Culture
- Spikiness – You have one superpower, and we don't care if you are weak in other areas
- Intellectual Honesty – You are radical about truth and candid to a fault
- Selective Intensity – You are an 80/20 person who gets the right things done
- Eternal Troubleshooting – You don’t tolerate brokenness, and can establish loops that identifies and eliminates it