Managing the Demands of a Client and Data Scientist for Optimal Product Results
In the dynamic world of data science and technical product development, collaboration is key. A recent project, managed under a co-manager model, demonstrates the power of this approach. The goal was to create device health management dashboards, showcasing high-risk devices requiring immediate attention.
The project team, comprising data scientists, product managers, and project managers, worked closely together. The client product teams were associated primarily with product managers or product owners, who served as the voice of the client within the team. These product managers ensured the product team delivered on the client's requirements, bridging the gap between client expectations and the development process.
The client's feedback played a crucial role in evaluating the results' accuracy and making necessary changes. By providing all this information, data scientists could empower clients to contribute to the project's success. They explained their data science processes in the clients' language, ensuring they understood how algorithms predicted events.
The client's active participation in all decisions and changes to the requirements, coupled with the team's commitment to transparency, led to the creation of a practical solution that created the expected business value for the analysts using the dashboards. The solution adhered to the Four Principles of Explainable Artificial Intelligence, including the principle of Explanation, which required providing accompanying evidence or reasons for all outputs.
The created dashboards showed details about how the models arrived at their conclusions, allowing analysts to click down to the underlying data for an explanation of the output. The UI of the different screens was easy to navigate and provided detailed explanations to analysts in their terminology.
Meetings were held weekly to discuss project status, next steps, and demos. Initially, these meetings were strained due to the client's dissatisfaction and the technical team's feeling of meeting their goals. However, as the client better understood the dashboard demos being presented and how the algorithms behind them were making decisions, the dynamic improved.
Better explainability and communication led to the client taking on more responsibility for the project's direction. This empowerment created a level of ownership for both teams, fostering a collaborative environment that ultimately led to the project's success.
In a previous role, the technical team worked with a client who understood data analysis but needed a technical team for machine learning and analytics. The co-manager model in project management, which assigns equal authority and responsibility to both the client and the technical team, proved instrumental in that project as well.
In conclusion, the co-manager model in project management, when combined with clear communication and client empowerment, can lead to successful data science projects that deliver practical, user-friendly solutions aligned with client objectives.
- The co-manager model, demonstrated in the project, leverages the expertise of both the client and the technical team in business, finance, and data-and-cloud-computing to ensure projects meet the client's objectives.
- As technology advances and businesses increasingly rely on data science and technical product development, the co-manager model in project management could become a key strategy in fostering successful careers within the industry.
