Success Factors for a Machine Learning Pilot

After running many successful machine learning (ML) pilots for organizations in various industries, we developed the following criteria to help us select the most suitable candidates for an early pilot with an organization. Since these early pilots are used to demonstrate the value of ML to achieve objectives that the organization cannot achieve using current approaches, they need to have a high probability of success. Of course, these criteria may also be used later to select ML projects among many potential candidates .

Scenario Selection Criteria for Pre-production Pilot

  1. Clearly defined Problem with clearly defined “Targeted insight” in mind
  2. Must have High (Business) Value for the organization
  3. Sharply defined scope that is time boxed to drive constraints
  4. Required Data – Available; Quality- Good; and Quantity- Sufficient to build Predictive Model
  5. Data Scientist(s) have direct, timely access to Domain specific Business and Technical SMEs
    • To clarify: the organization must be willing to make these resource(s) freely available for duration of pilot
  6. Clear sponsor who can make fast decisions and unblock obstacles
  7. Technical Resources with Expertise: the organization needs to provide experienced technical resource to acquire data from internal or external data sources
  8. Significance of dependencies, challenges, and roadblocks
  9. [Can be operationalized with existing products (optional)]

We also have developed a scoring model so the organization can objectively compare several potential projects according to the above criteria.

Follow My Blog

Get new content delivered directly to your inbox.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s