Our collaborative data projects are 12 weeks in duration and give you access to a wide range of expertise across our team of data scientists and engineers. We work alongside your team to scope your data science or engineering project, build a prototype solution, and deploy it within your organisation. Data science solutions can include models, reports, dashboards. Data engineering solutions can include data warehouses, data pipelines, and data applications.

We usually spend 2 hours per week face-to-face with your team, either virtually or in-person. Collaborative working is encouraged so the more time that your team is able to spend on the project the better. Upskilling is a key part of collaborative projects, so any opportunities for learning will be taken advantage of, and we are able to provide bespoke tutorials on areas relevant to your project.

Projects begin with business understanding and project scoping. We encourage you to clearly define a goal which can deliver value to your organisation. Specific business objectives are then built on this goal. Finally, we define possible data solutions which can meet your objectives. We use a scoping document called a project charter to bring all of these elements together and provide clear documentation which can be shared to stakeholders within your organisation.

The next phase is data understanding, where we dig deep into your data and explore its structure, patterns, and trends. Data cleaning, which involves handling missing or erroneous data, can be a big part of this phase. After the data has been cleaned, we will work to bring it into the correct shape for the data solution defined in the project charter, through a process called ‘data wrangling’. If your project is more focussed on collecting data, rather than working with data you already have, this phase can be used to define an appropriate structure for data collection.

Once the data has been appropriately cleaned and reshaped (wrangled), we will start building your data solution. This is an iterative phase where we will design a prototype, review it, gather feedback, and work to improve it. If the solution is a model, a measure of accuracy is usually used to assess the quality of the prototype. If the solution is a report, dashboard, or other data application, more qualitative feedback will be required from users of the solution.

The final step in the process is production. This is where we help you to integrate the prototype into your existing processes or make it available to your customers. Monitoring and maintaining the solution is a key part of this step and we will explore approaches to this with you. Our team has expertise in cloud computing so we will often recommend this for production solutions, however we are also able to help you deploy on-premise where necessary.

Your project may or may not get as far as production. There is always an element of risk involved in data science and engineering projects, therefore we may only get as far as the prototyping stage. This is not always a bad thing and the learning that comes from prototype development is often invaluable to organisations in the early stages of their data journey

Throughout the project, we will keep key stakeholders updated on what has been done and learned by both our team and yours. We encourage you to organise review sessions at key milestones to communicate successes to your wider team and gather feedback on the direction that the project is taking. Review sessions provide a valuable opportunity to reflect and reprioritise. This may involve a significant change in direction or approach where necessary.

At the end of the project, we will write a report on what we have successfully done and learned. We welcome any input from your team into this process. The final report, once approved by your organisation, will be made available to the public. This provides an opportunity for you to share the exciting progress you have made with the world.

• SME (size and turnover limitations?)
• Within state aid limitations?
• Able to share data (this can be anonymised or synthetic sample data)
• Consent to share case study report publicly