Our engagements are conducted according to a well defined three-phase approach.
Identify and Execute a Pilot Project
The project should be limited in scope, complexity and duration (less than three months), so results can be shown and value can be realized quickly.
Define AI Strategy and Roadmap
A set of broader goals are identified, a strategy is put in place to execute towards these goals. This phase includes architecture and technical considerations, labor resource needs, budgets and milestones. The outcome is a set of projects to be executed over the next 6 to 24 months
Execute Projects
The projects identified in the previous phase are executed according to our methodology.
Our methodology is based on CRISP-DM. It is adapted for a highly agile implementation and greater enablement of continuous improvement.
| Business Understanding: Define and understand business requirements and problem to be addressed. |
| Data Understanding: Identify and understand available data, select analytical approach to be followed to address the problem. |
| Data Preparation: Clean and transform data, create derived data that will be consumed by the model. |
| Modeling: Select, build, train and assess model. |
| Evaluation: Assess quality of the model against business objectives and ensure it is working as intended. |
| Deployment: Roll out the solution to end users. |
| Monitoring: Monitor the solution performance and impact, incorporate users feedback, make adjustments for as long as the solution is used. |