Our focus is on the more cutting edge areas of machine learning like Active Inference /First-Principles AI and Probabilistic Programming. We also do more traditional approaches.
Data cleaning and imputation using traditional as well as Bayesian tools.
We do regression and classification using traditional as well as Bayesian approaches.
We do dimensionality reduction and clustering using traditional as well as Bayesian approaches.
We do Reinforcement Learning and Active Inference.
We do Probabilistic Programming using tools like RxInfer, Pyro, and NumPyro.
We do Active Inference, based on the latest in neuro-science.
We do interfacing with a variety of cloud providers.
Depending on the problem, we apply the most appropriate approach.
Kobus is an excellent data scientist and top-tier expert that has provided my organization with excellent support to my direct program customers. His expertise was instrumental in the future of providing supply forecasting capability as an analytics service to the customer. Kobus' ability to immediately adapt to a new team through a virtual workspace, listen to our initial ideas while evaluating our problem set, and rapidly return with a concept for how to approach and potentially solve the prediction problem was quite impressive and let us know immediately we had a team member that was up to the task at hand and he exceeded expectations with his expertise and teamwork. I was impressed by his ability to not only develop an approach to what was a very complicated problem, but as well the ability to document the approach, process, and outcomes in detail and with the provided breakdown for novices to data science to easily understand and explain to our customer leadership. A real asset to any data science team and expert problem solver I would recommend to any team.
Kobus is a wonderfully disciplined data scientist who is fantastic to work with. He carefully digs into the requirements of a problem and thoroughly analyzes different solution approaches. He has a wide range of software skills at his fingertips and loves to engage with his colleagues in exploring how new technologies and approaches can be applied to problem solving and visualization. I look forward to any future opportunities to work with Kobus as he is the type of person that really grows on you in a team setting. As a peer, you look forward to the next lucid technical conversation. As a project manager you enjoy the predictability and confident deliverables that stem from his abilities.
I had the pleasure of working with Kobus at the Boeing Concept Center. He was proficient in a variety of programming languages in addition to being able to deep-dive across technical data science. Considering the amount of IoT sensor and telemetry data ingress we were handling in constrained airborne environments at the time, it was great to see insights through all that data. I would recommend Kobus in a variety of roles specifically data science and enterprise applications. I would work with Kobus again.
Kobus has great architectural level of the software development expertise as well as the ability to define analytical approaches in solving software quality problems. His deep knowledge of theory of control and algorithms allows him to formulate interesting directions in predicting dynamic of a software system parameters and its control. He is opened to share his knowledge with his colleges. Kobus is continuously learning new software directions and applying high level of logical analysis to select candidates for an existing software system improvement. Developing and implementing software simulation plan in order to meet any business requirements, also define Kobus as a knowledgeable specialist. His is a great asset for any company that targeting high level of development and it is a pleasure to work with him as a person.
Kobus is an excellent software engineer and operations researcher. He did ground-breaking work in event based simulations. I will work with him to solve complex problems any day, especially in solving complex modelling and simulation systems.