How to effectively engage business domain experts in building Data & AI products

Three-level Engagement Framework

Sandeep Uttamchandani
4 min readFeb 21, 2022


Building Data & AI products is a team sport: application data producers, data engineers, analysts, scientists, ML engineers, DataOps, MLOps, ITOps, full-stack developers, product managers, release managers, and business domain experts (such as sales, product, marketing, support).

Image Credit:

The entire cross-functional product team needs to work in unison to:

  • Build the right product (including GTM)
  • Building the product right
  • Deliver seamless and robust operational experience to the users

Among the personas listed above, teams often do not prioritize engaging the right business domain experts. Simplify having a business owner who is funding the initiative is required but not sufficient. Most often, the goals of engaging the business domain expert are not well-defined and often fuzzy.

Suppose you are building an internal AI product for churn prediction — the product will be used by sales agents to proactively identify customers at risk and take mitigation actions to reduce the overall churn. For this product to be successful, the business domain expert (in this example, an operational leader or architect from the customer support team) needs to be intimately familiar to guide the team to pick the right data, build the right insights from the data, and deliver the insights in the right form to truly deliver business impact.

The business domain expert needs to be engaged at three levels: Design thinking, Data thinking, and Operational thinking. It is seldom a single person, and more often a team of 2–3 business domain experts to effectively engage at all the levels.

Design thinking



Sandeep Uttamchandani

Sharing 20+ years of real-world exec experience leading Data, Analytics, AI & SW Products. O’Reilly book author. Founder #Mentor #Advise