Improbable’s SKYRAL platform enables government and military users to respond to a diverse range of complex threats in a flexible, tightly integrated way. Built from the ground up with open standards and supported by an ecosystem of partners from across industry and academia, it’s a more efficient, flexible and cost-effective way to build and operate synthetic solutions.
Model Engineers at Improbable Defence are focussed on creating the models that will power the next generation of decision support technology. We work closely with applied scientists and software engineers to create these models and the tooling to support them. Our culture is friendly, relaxed and inclusive. We divide up ownership across the team, so expect responsibility early on, but the team is large enough that you will be fully supported by colleagues.
Our primary responsibility is to deliver robust, performant, scalable scientific models into production that run in simulation and sit at the heart of our core product offering. This means building data pipelines to generate accurate representations of real-world environments and the algorithms for simulating their dynamics. We also build tooling that enables us to develop, test and integrate models faster. We do this on the timescale of days and weeks, not months and years. We are not a Product Development team; we are a customer-facing, solution-focussed Rapid Application Development (RAD) team.
- Productionising prototype models ensuring they are scalable, robust and performant.
- Developing and implementing algorithms which result in plausible modelling behaviour.
- Building pipelines to extract, synthesise and integrate data from various sources, making it available to simulation engines and user interfaces.
- Building tooling that enables model developers to do their work quicker and better.
- Profiling and improving the performance of models so that we can reach new levels of speed and scale.
- Identifying and implementing generalisations and abstractions of existing models so that new models can be created easily and efficiently.
- Contribute to building a next-generation product which will help governments gain a richer understanding of their most critical problems through the power of virtual worlds.
- You will work closely with applied scientists to take prototypes and model designs into a production environment. The main languages we use are Python and C++.
- You will support other software engineers and applied scientists in developing best practices.
Why You're Made For This:
- Significant experience working as a professional engineer, applying model and data engineering to solve customer problems.
- Skilled in Python including the more advanced features of the language (e.g. decorators, and built-in methods).
- Professional experience of tools and libraries for modelling and scientific computing (e.g. NumPy, SciPy, PyTorch, Scikit-learn)
- Experienced with data manipulation and processing, using technologies such as SQL or pandas
- Ability to think holistically about every aspect of software development, including testing, documentation, security and performance.
- Experienced in working on large projects alongside product managers, architects, delivery managers and quality engineers
- History of working closely and effectively with external customers, exploring their problem space, understanding their requirements and building solutions to satisfy them (preferred).
- Track record of sharing knowledge with and learning from others, and actively mentoring colleagues to improve their skills.
- Experience with C++ is a benefit but not essential.
We asked employees of Improbable how satisfied they were with flexible working, and this is what they told us
Working at Improbable
Spitalfields, London (next to Liverpool Street Station)