Senior AI researcher
This is a full-time position based in Prague.
Rossum is a fast-growing Artificial Intelligence company that builds technology for understanding documents. Every single day, more than 100 human lifetimes are spent worldwide retyping information contained in invoices and we think that’s completely crazy. So we are now working to liberate millions of people from similar boring and repetitive work.
Rossum’s product is number #1 in its category. At the same time, at Rossum, we do not just repackage existing research. Our technology is instead based on original work meshing computer vision with natural language processing in a unique way using deep learning. Our research team is in a unique spot – we are treading in an area with little prior research where we need to keep inventing original approaches, yet the work is motivated by immediate applications. We then train our models at scale and production rollout is always just a single merge away.
What you will do
- Help us solve the challenges involved in making computers understand documents
- Work on open research problems associated with information extraction from layouted documents. (Moving from a strongly supervised pipeline to an end-to-end weakly supervised setting. Truly fusing NLP and computer vision approaches. Exploring process learning over task learning (iterated extraction, reinforcement learning), one-shot learning approaches. And many more.)
- Be ready to supervise a small team of researchers and programmers working with you, collaborate with other senior researchers.
- Follow the literature and co-lead our research meetings and reading group.
- Over time, present Rossum’s research – mainly at academic venues, but also professional conferences, our blog, etc.
- Write quality code and participate in code reviews. Even if programming isn’t your main focus, your code will run in production.
What we expect
- Great theoretical background in machine learning principles.
- Track record in academic publications or equivalent (top Kaggle, open source, etc.) that proves deep learning experience.
- A good overview of the latest progress in deep learning.
- Some experience with GANs, Transformers and/or RL a plus.
- Some Python programming experience, Keras/tf experience a plus.
- Basic working knowledge of Git, the UNIX command line (Linux or Mac) and *TeX.
- Passion for extending the boundaries of machine learning, and the courage to apply it to even daunting tasks.
What we offer
- Energy-charged environment of a tech-first startup focused at original AI research, with a rapidly scaling dataset, in a unique product-driven setting: It’s not unapplied fundamental work, but it’s far from just reusing published ideas either.
- Work and swap ideas in a research team that works in state-of-art deep learning. Our ambition is to grow into one of the most advanced commercial machine learning research teams in Europe. What about your ambition?
- Enjoyable working environment: A diverse international team of geeks, globetrotters, artists, and health enthusiasts in the heart of Karlin. We offer flexible working hours and highly value independence. Perks like tasty and nutritious snacks and the best coffee and tea in town are a given.
- Independence and flexible working hours: Some of us come to work at 7:30, some of us at 11:30 and it is up to each of us to organize our daily agenda and get the job done.
- Opportunity to attend machine learning conferences.
- Potential Silicon Valley / Stanford research residency opportunity in the future.