Think of us as your data science department

Today, data drives everything and we love driving data. We want to help you turn your data into business insights. We are the data science team you didn't have to hire yourself.

Our Story

Data Science for those that don't do Data Science

Sweating data is hard work. The majority of companies do not have the skills or the frameworks to get the most out of their data. Why should they? After all, their focus is on building their business, not poring over data frames.

We formed ModelInsight to give companies the power to explore and gain insight from their business data. We had a theory that companies would do more in this field if there was a company like ModelInsight to help them.

Our offer

On-demand Brain Power

We are a "turn on-off-able" team. Like the cloud, you only pay for what you use; no retainers, no big up-front fees, no sign-up costs. Get the data analysis capabilities that your competitors have but at a fraction of the cost.

We specialise in the value-add end of data science and have expertise in Machine Learning, Big Data, Search, Natural Language Processing, Statistical Modelling, and Data Visualisation. We can also help with data strategies, architectures and data cleansing techniques.

How we work

We assist with all steps in the data science work flow

  • Acquire
  • Prepare
  • Analyze
  • Visualise
  • Share

Talk to us first. We offer an initial data problem assessment with no obligation and no fee.

Contact us today. You supply the coffee and we supply the science.


We are offering some top training sessions in April. You should also sign up to our low volume email announce list to hear about new courses.

2 day High Performance Python course

Learn high performance techniques for performant computing, a mix of background theory and lots of hands-on pragmatic exercises


  • Profiling (CPU, RAM) to understand bottlenecks
  • Compilers and JITs (Cython, Numba, Pythran, PyPy) to pragmatically run code faster
  • Learn R&D and engineering approaches to efficient development
  • Multicore and clusters (multiprocessing, IPython parallel) for scaling
  • Debugging strategies, numpy techniques, lowering memory usage, storage engines
  • Learn Anaconda and conda environments to simplify your scientific coding
  • Learn from our years of hard-won experience so you can confidently develop performant Python code

Aimed at: Intermediate Pythonistas (you need higher performance for your Python code)

Central London, Thursday 9 April to Friday 10th April 2015

£700 (two full days, includes lunch)

Click to Book

2 days - Understand Statistics and Big Data using Scikit-Learn and Friends

Learn everyday statistics and start machine learning

Day 1 goals:

  • Probability vs statistics. Understand distributions and the difference between frequentist and bayesian approaches.
  • Learn the pitfalls of experimental design that lead to invalid inference. Demystify p values (and learn why they're important).
  • Learn linear and logistic regression and why they are fundamental tools for making sense of data.
  • Learn how to apply these lessons in python. We'll draw some valid conclusions and, more important, some invalid conclusions to help you recognize bad experimental design.
  • Use pandas and statsmodels to make sense of your data.
  • Use matplotlib and seaborn to visualize data sets.

Day 2 goals:

  • Use scikit learn and statsmodels learn to understand bayesian spam detection.
  • Use scikit learn and statsmodels to understand quality testing and insurance.
  • Learn how to use Google trends to probe huge data sets in interesting and fascinating ways.
  • Do you measure your own life? Then you'll like this practical example: use scikit-learn to understand our own behavior.

Aimed at: Beginner/intermediate Pythonistas with some mathematical background

Central London, Tuesday 7 to Wednesday 8 April 2015

£700 (two full days, includes lunch)

Click to Book

Our Clients

We've worked with leading brands and small to medium businesses

Royal Mail

Evolutionary Algorithm based Planning System to create fast adapting logistics schedules for delivery routes

You Gov

Realtime Brand Measurement on Twitter and Facebook using Sentiment Analysis and Named Entity Recognition

First New York

Live Hedge fund trading strategies developed using statistical models

Arctic Shores

Statistical verification of data modelling approaches

Kent County Council

Establishing a strategy roadmap for big data and analysis


Building models of commuter behaviour using timeseries analysis to understand the needs of urban travellers and the transit system's behaviour


Automating contract job and contractor matching to enhance efficiency of the industry using Machine Learning, Natural Language Processing, Python, C# and Big Data

Desktop Genetics

Providing high performance DNA search algorithms to enable client's AutoClone to solve real-time search requirements

Our Team

Pragmatic data scientists

  • Ian Ozsvald

    Principal Pythonista

    Data Scientist and Artificial Intelligence Researcher, Ian works to understand business objectives to then deliver IP for his clients. Author of O'Reilly's High Performance Python, holds a Masters Degree in Artificial Intelligence. Skills include machine learning, natural language processing and recommendation systems in Python.

  • Chris Brown

    Data Dude

    A technologist with industry-proven management skills for front to back solutions delivery. Specialises in requirements capture and solutions architecture for data platforms. Software development, predominantly in Java-based applications. Holds a Masters Degree in Financial Engineering. Current research areas are big data architectures on cloud provisioned environments and agile approaches to data science problems.

  • Bart Baddeley

    Statistical wizard

    Backed by a PhD in Artificial Intelligence Bart predicts and models large, noisy datasets using Statistical Modelling and Machine Learning using Python and Matlab. Specialisms include Recommender Systems, eCommerce, computer vision and physics.

Contact Us

Call us and talk data

Tel:+44 (0) 7970 405682

We are open Monday to Friday from 8:00 am to 7:00 pm GMT (United Kingdom)