What you’ll be doing:
You will be a core contributor to our core machine learning models.
Modern NLP models that solve intent recognition, entity extraction
against vernacular text. These models have to scalable, available and
secure. This is critical to our success. You’ll become a core member of
the engineering team.
We’re dedicated to what we’re doing and we work hard in a sustainable
and healthy way. If you want to work from home some days that’s cool.
All that matters is you contribute effectively and work well with the
team.
Skills and experience
Sorted from “essential” to “nice-to-have”:
Read machine learning literature and implement the models described (eg from recent papers in ICML/ICLR/NIPS).
Experience building and testing models.
Comfortable with core concepts from linear algebra, probability and statistics required in ML.
Python.
Deep Learning frameworks eg Theano, TensorFlow, Torch.
Working knowledge of GNU/Linux
An undergraduate degree in a hard science, mathematics or signal
processing, engineering subject and a postgrad degree or PhD in ML.
Solving Natural Language Processing problems.
Have used "Infrastructure as a Service" (e.g. Google Cloud, AWS)
Containers (Docker, CoreOS etc.) and orchestration (e.g. Kubernetes/Mesos)
Distributed data storage (e.g. RethinkDB, ElasticSearch) and SQL
Stack:
The backend is a mixture of Python and Rust and consists of many
containerized stateless microservices and a few stateful ones. There’s a
mixture of interactive and batch services, with many running heavy
numerical loads. Everything gets deployed on Kubernetes in Google Cloud
Platform. We use RethinkDB and Elasticsearch for storage and indexing.
The frontend is built using Typescript, React, Redux and Webpack.
Team:
re:infer was founded by Machine Learning researchers, engineers and
entrepreneurs from UCL and Google. We’re a small team of technologists
(~ 12 people) motivated by solving the very hard problem of getting
machines to understand conversations.
We work in two-week sprints following the best bits of the Agile
approach to software development. Our employees benefit from a relaxed,
social atmosphere in which they’re free to make use of our open-plan
office to discuss ideas, or put their headphones in and find a quiet
corner as and when they require. We support flexible working hours and
have a generous holiday policy and company pension. Everyone gets their
own top-spec laptop and whatever office hardware works for them.
Business:
re:infer is an artificial intelligence solution that turns
unstructured communications data (emails, chats, calls) into structured
actionable data.
re:infer is a software solution that can read and understand human
communications data found in emails, calls and chat. It self-learns, can
classify intents, emotions and entities in every communications and
converts them into structured data that provides our clients with an
automated MI capture and can drive downstream automations.
Our Enterprise Communications and Analytics platform is deployed
globally and across multiple industries to transform our client’s
operating models and deliver scalable and transformational automation
initiatives
re:infer is one of the most exciting A.I. startups in Europe. A
spin-out from the world famous AI research group at University College
London that produced SearchSpace and DeepMind. We have won ‘Best AI
startup’ at the World AI Summit and are regularly featured in the top
lists of breakout machine learning companies operating in Europe. The
founders are leading researchers in the field Machine Learning.
re:infer has great investors like IP Group, Crane Ventures and we’re
backed by some of the world’s leading AI entrepreneurs. In February we
closed our seed round Seed Round of $3.5m. So you are joining us at the
best possible time. Come and build the future with us. This will be the
best work of your life.
re:infer’s offices are at Mindspace Aldgate, in the heart of East London.