Forecasting ML Model: 4-Week implementation

Artificial intelligence (AI) and machine learning (ML) have been key focus areas for us. Our 4-week implementation will provide an overview of Microsoft data analytics and ML platform. We will evaluate your current business needs, help you decide the right Azure resources, and analyze key metrics to develop an ML model.

We will suggest Azure resources based on your scenario. The ML model will be built on the Azure Databricks platform using Python. We will showcase insights with Microsoft Power BI reports.

Target Audience:

• Business Strategists
• Data Scientists
• IT Business Heads

Activities:

Week 1

Analysis

• Understand business and forecasting requirements
• Understand key metrics and any dependent parameters and data points
• Study business domain and perform industry research on key metrics
• Understand the current architecture to suggest the best Azure service for this engagement

Week 2

Platform Setup

• Study current data ingestion methods, sources, and storage methods
• Work with your team to set up a data ingestion and storage platform
• Onboard data with all essential metrics and dependent variables
• Cleanse data and build the dataset

Week 3

Metric Analysis

• Perform seasonality analysis and dependent variable analysis of key metrics
• Provide observations and key inferences of analysis

Week 4

Modeling

• Build and train the model using various combinations of data
• Visualize the data and verify the results of the model build

Visualization and Roadmap

• Create a roadmap based on the data and insights generated by the model
• Build 1–2 Power BI views using the key generated insights

Deliverables:

• Overview of Azure Data & AI with Power BI capabilities
• Setup of Azure platform to onboard data
• Forecasting model
• Insights on key metrics
• 1-2 Power BI views

Outcome:

• ML Solution with Azure Cloud architecture that accurately forecasts key business metrics