Foundation for AI in the Enterprise

At our January roundtable, together with Daria Hvížďalová — an expert in AI implementation who leads executive AI programs at MIT (Massachusetts Institute of Technology) — and Petr Svoboda, CEO of CodeNOW, we discussed what the foundations of AI adoption in large enterprises should look like — and where companies most often fail, especially when moving from pilot to production.

Key Discussion Points

The need for “uniqueness” in corporations
Many companies feel the need to build their own platforms and tools instead of commoditizing strategic components. This often extends the time between pilot projects and full production rollout.

Daria’s recommendation:
Invest in internal education and develop subject matter experts as ambassadors. Real change gains momentum when employees understand how AI impacts their daily work.

Petr’s recommendation:
Break large initiatives into smaller domains. Assign each part a mini-budget and a clear owner.

Security and Governance

Clear rules and auditability are essential.
Companies often begin with experimentation and only later define policies. It is critical to introduce security and governance rules in parallel from the very beginning.

Commoditization vs. Uniqueness

Many infrastructure solutions can and should be commoditized.
Reserve internal investment for areas that truly create competitive advantage.

Practical Recommendations for Management

  • Assign a clear accountable owner to each pilot.

  • Approve small pilot budgets with defined success metrics.

  • Build a network of internal ambassadors to spread knowledge.

  • Carefully decide what can run as SaaS and what must remain on-premises.

  • Measure impact using real business metrics: time saved, cost reduction, process optimization, and error rate reduction.