How We Work


We start by understanding your problems and opportunities better. We’ll investigate the value of using machine learning and how we can best partner with you. Working with you, we'll develop a roadmap for subsequent work.

Data Preparation

Developing machine learning models requires the collection and augmentation of appropriate datasets. We will work with you on the data you have and search for external datasets. We clean and structure this data in preparation for machine learning development.

Model Development

Our experienced engineers and data scientists design, build, implement and deploy machine learning models for your product/service. Typically this will be an iterative process, starting with a MVP (minimum viable product) and then later developing a better market-ready product. We will work with you to test the model, analyse the results and then make further improvements.

How we work together

Every client is unique but typically we work together as follows:

  1. Initial exploratory meeting of 90 minutes.

  2. Scoping: a two to five day project to determine and derisk potential projects.  

  3. Initial project: a fixed price project with agreed deliverables, with all IP assigned over to you.

  4. Follow-on projects on similar terms


“We have developed a great partnership with DataJavelin, having now worked with them on several projects. Their expertise in machine learning and the ability to develop practical solutions is well appreciated by us. Working with them we have been able to develop AI-powered algorithms and more intelligent forecasting systems to improve accuracy and reduce costs relative to traditional approaches.” 

David Martin, CTO Ambiental