We are a bespoke research and development studio and data consultancy based in Cambridge, UK. We help companies find solutions to complex problems in the domains of algorithms and data.
We are computer scientists and engineers with background in mathematics, artificial intelligence, and operations research. We have experience in applying algorithmic methods across a variety of fields. We also have a track record of creative innovation in problem solving, both in academia and industry.How we can help?
We provide expertise in algorithms, data structures, data analysis, machine learning, artificial intelligence, optimization, mathematical modeling, complex networks, and data compression. Our customers often come to us for solutions when they do not have the necessary skills and expertise in-house.How we work?
We analyze your problem, conduct research, experiment with models, design a prototype, propose a solution, and deliver an efficient implementation. We work along your scientists and engineers. We have successfully worked with a variety of customers from early-stage startups to R&D departments of large corporations.
We design, develop, and implement bespoke algorithmic solutions to complex problems. Our customers are businesses that need a novel approach, an innovative idea, or expert knowledge.
We build on years of algorithmic research and connect science to engineering practice. We have worked with companies from different sectors – 3D graphics, gaming, mobile communication, streaming, artificial intelligence, R&D.
Mathematical optimization allows to find the best solutions to problems. We have experience in mathematical modeling, linear and integer programming, scheduling and routing problems. We have developed exact and approximate algorithms as well as designed heuristics for several discrete optimization problems. We are currently working on an internal R&D project to bring machine learning to combinatorial optimization.
Custom data requires custom codecs. A dedicated compressor will deliver a better compression ratio and encoding/decoding times than generic solutions. We have developed lossless and lossy compression engines for images, time series, 3D point clouds, and computer graphics data. We also have experience in designing metrics to measure the quality of lossy reconstruction. We are currently experimenting with codecs that are not engineered but learned from data.
We particularly enjoy working with graphs and networks. (Hence our name.) Graphs are ubiquitous data models and are extensively used in several domains. We have experience in complex network analysis, development of graph algorithms, probabilistic graphical models, graph signal processing, and graph neural networks.
We help companies leverage their data. We look for patterns and insights, build predictive systems, and enjoy designing custom algorithms for users' data. We have experience working with all kinds of data sets – from images and 3D data to brain signals to financial data.
We analyze data to uncover patterns, insights, and interesting information hidden in your data sets. We start with exploratory data analysis and use statistical and machine learning techniques. We look for questions that can be answered with customer's data set and help companies discover the value of their data. We also advise on how to collect data and how to design an experiment.
We build predictive models. We have experience in forecasting time series, estimating quality of image reconstruction, and predicting links in graphs. We have worked on classification problems in a number of domains. We can deliver a proof of concept, search for the best model, and build a complete machine learning system.
We have experience with a range of machine learning techniques – from Bayesian inference to graph neural networks and reinforcement learning. We have also worked on non-standard applications of machine learning to data compression and optimization.
We specialize in all things data and algorithms. We provide scientific and engineering consultancy in the areas of computer science, data analysis, and software development. We assist various stakeholders at your organization – managers, data scientists, software developers, researchers – providing answers in our area of expertise.
Contact us to see how we can work together!
We conduct research on behalf of our customers. We look for robust solutions to your technical problems in applied mathematics and computer science. We design new algorithmic techniques and provide answers and insights. We also run a number of internal R&D projects and conduct academic research.
Contact us if you face a challenging technical problem!
We provide software development services, usually as part of our consulting or R&D engagement.
Contact us to discuss a software product you need!
We also offer custom-tailored training for teams and individuals in algorithms, data analysis, software development, and artificial intelligence. We have years of academic classroom experience.
Contact us if you require a training course!
4colors Research was founded
in 2016 by Marcin Kamiński. We are based in Cambridge, UK but we work with customers around the world.
Our roots are in science and we enjoy playing with, combining together, and discovering new ideas. We like the challenge of finding a good solution.
We are a small team of scientists and engineers but we maintain a network of collaborators. We work closely with our clients on solving their problems.