How Commerce Leaders Should Think About AI and Data Platforms
Ashish Kasama

How Commerce Leaders Should Think About AI and Data Platforms

Over the past few articles, we’ve explored why commerce analytics breaks at scale, why unified Lakehouse architectures are becoming the standard, how a modern commerce intelligence architecture is designed, and which AI use cases create real business impact.

The final and most important question for commerce leaders is not what AI can do—but how to adopt it responsibly, effectively, and at scale.

This article outlines a practical framework for how commerce leaders should think about AI and data platforms in today’s fast-moving, transaction-driven environment.

1. Start with Decisions, Not Technology

One of the most common mistakes organizations make is starting their AI journey with tools.

AI initiatives should not begin with:

  • Model selection

  • Platform comparisons

  • Feature checklists

They should begin with decisions:

  • Which commerce decisions most directly impact revenue or margin?

  • Where does volatility hurt the business most?

  • Which processes are still reactive or manual?

Examples:

  • Replenishment decisions

  • Pricing and promotion planning

  • Fraud and returns management

  • Personalization and growth optimization

AI should be introduced where decisions matter—not where experimentation is easiest.

2. Treat Transactions as Strategic Assets

Commerce organizations run on transactions.

Orders, payments, returns, inventory movements, and customer interactions are not just operational records—they are the raw material for intelligence.

Leaders should ensure:

  • Transactional data is captured comprehensively

  • Data is available with minimal latency

  • Analytics and AI teams work directly on transactional datasets

When transactional data is delayed, fragmented, or abstracted away, AI initiatives lose relevance.

3. Choose Platforms That Unite Analytics and AI

A critical platform decision for commerce leaders is whether analytics and AI operate:

  • In isolation

  • Or on a shared foundation

Fragmented platforms increase:

  • Time to insight

  • Cost of ownership

  • Operational complexity

Unified platforms enable:

  • Faster experimentation

  • Easier production deployment

  • Consistent governance

  • Lower long-term risk

Commerce leaders should prioritize architectures that bring data engineering, analytics, and AI together, rather than stitching them together downstream.

4. Scale AI Incrementally, Not All at Once

Successful commerce organizations do not attempt to “AI everything” at once.

Instead, they:

  • Start with high-impact use cases

  • Prove value quickly

  • Expand systematically

A common progression:

  1. Forecasting and inventory intelligence

  2. Pricing and promotion optimization

  3. Personalization and growth

  4. Risk and fraud intelligence

This phased approach reduces risk while building organizational confidence in AI-driven decisions.

5. Embed AI into Workflows, Not Dashboards

AI delivers value only when it influences action.

Leaders should ensure AI outputs:

  • Flow into pricing systems

  • Feed inventory planning tools

  • Trigger alerts and recommendations

  • Support operational teams directly

AI that exists only in dashboards or reports remains underutilized.

The goal is not visibility—it is execution.

6. Govern for Trust, Not Control

Trust is essential for AI adoption in commerce.

Leaders should focus on:

  • Data quality and lineage

  • Model explainability

  • Clear ownership of metrics

  • Controlled experimentation

Overly restrictive governance slows innovation.
Too little governance erodes trust.

Modern platforms allow governance to be applied consistently across data, analytics, and AI—without slowing teams down.

7. Build Capabilities, Not Just Solutions

AI in commerce is not a one-time project.

Leaders should invest in:

  • Reusable data foundations

  • Shared feature sets

  • Scalable ML pipelines

  • Cross-functional collaboration

This enables the organization to:

  • Add new use cases faster

  • Adapt to market changes

  • Avoid constant re-platforming

Capabilities compound over time.
Point solutions do not.

Final Thoughts

AI is rapidly becoming a core component of modern commerce strategy.

But success does not come from adopting the most advanced models or the latest tools.

It comes from:

  • Clear decision focus

  • Unified data and AI foundations

  • Phased, outcome-driven adoption

  • Strong governance and trust

Commerce leaders who approach AI as a platform capability, rather than a collection of experiments, are best positioned to compete in an increasingly volatile and data-driven market.

The future of commerce belongs to organizations that can learn from every transaction—and act on it in time.