Member since October 2026
Karina Mikucka helps leaders make sense of AI and decide how to put it to work. She brings 20 years of enterprise technology leadership, including eight years leading an AI and digital function in regulated life sciences.
Her talks connect enterprise leadership experience with hands-on work building AI agents. She explores how organizations move from AI usage to meaningful outcomes, establish governance that enables responsible use, and design workflows where people and agents have clear responsibilities, authority, and accountability.
Previously, Karina served as Head of Digital Business Solutions at Foundation Medicine, a Roche company, with leadership responsibilities spanning enterprise AI strategy and enablement, governance, platform strategy, and digital transformation. Earlier, she built the product and project management function at New York Road Runners.
Karina holds a master’s degree in Information and Knowledge Strategy from Columbia University and bachelor’s degrees in Applied Mathematics and Statistics and Economics from Stony Brook University.
Based in Boston, she offers keynotes, leadership workshops, strategic briefings, and panel discussions, virtually and in person. Every session is tailored to the audience and ends with something participants can put into practice.
Speaking topics
• Enterprise AI strategy and transformation
• Responsible AI governance
• Leadership and decision-making at AI speed
• Human and AI agent collaboration
• Efficiency AI vs Opportunity AI
• Agentic AI operating models
• Measuring AI business value
• AI literacy, adoption, and organizational change
• Women building with AI
I love the moment when AI becomes something people can question, shape, and build with. Helping someone move from uncertainty to trying something themselves is especially rewarding. I care about how we use that confidence: to improve work, make better decisions, and give more people a voice in shaping what comes next.
Generative AI was spreading through a regulated business, but no single function owned the enterprise risk. I brought three options to senior leadership. We chose to broaden ownership so the functions affected by AI decisions helped make them.
A cross-functional task force developed the responsible AI policy with no incremental headcount. The CEO approved it, and it became the governance foundation for the enterprise Copilot deployment and subsequent AI use cases.
The lesson I share from the stage: governance is an accelerator - it can help an organization move forward when people understand what they can do, who decides, and when to ask for help.
My path has taken me through mathematics, economics, product management, and enterprise technology leadership. Across those disciplines, my focus has been on connecting technology decisions to how organizations work and what they want to achieve.
After eight years leading an AI and digital function in regulated life sciences, I wanted to explore firsthand what changes when AI starts carrying out work. I spent eight weeks building the agent system I now use in my own work.
Building brought leadership questions into sharp focus: What should an agent be allowed to do? How do we assess its work? When must it stop and hand back to a person? Those questions now shape my advisory work, talks, and workshops.
Talk, panel or fireside · 30 to 45 min
AI can help organizations do existing work faster. It can also make possible work they could not previously do: serve an overlooked audience, offer a new service, or pursue an idea that once cost too much to explore. Karina examines how leaders distinguish efficiency gains from new opportunities, question the assumptions behind today’s workflows, and decide where to experiment.
What they leave with: a practical lens for distinguishing efficiency and opportunity use cases, questions to uncover possibilities beyond the current workflow, and a way to test whether an idea creates value worth pursuing.
A strong fit for leadership teams, innovation audiences, and moderated conversations about AI strategy and growth.
Talk, panel or fireside · 30 to 45 min
As AI expands what teams can produce, leaders must become more deliberate about what deserves to be done. Karina explores how work changes when people and agents share responsibilities.
What they leave with: questions to ask when redesigning organizational ways of working. Top leadership capabilities that take your leadership to meet the moment.
One of her most adaptable session, and the strongest fit for panels and moderated conversation.
Talk, panel or fireside · 30 to 45 min
When AI starts carrying out work, leaders must decide how to divide responsibilities, grant authority, assess quality, and handle exceptions. Karina explores how work changes when people and agents share a workflow, and what must be in place before an agent is trusted to act.
What they leave with: a test for which work belongs to an assistant and which belongs to an agent, and the conditions that have to hold before an agent is allowed to act.
Her most adaptable session, and the strongest fit for panels and moderated conversation.
A hands-on AI agent workshop for leaders
Workshop · Half day, or 4 × 90 min virtual
Leaders build a bounded agent prototype during the session, then test it against what decides whether an agent can be trusted in a real workflow: explicit permissions, quality criteria, and escalation rules for when it must stop and hand back to a person. Karina spent eight weeks building the agent system that now runs her own work, and the workshop distills what that taught her about the components of one and its organizational impact.
What they leave with: a working prototype they built themselves, the permission, quality and escalation rules they tested it against, and a clear-eyed list of what their own organization is not yet ready to let an agent do.
For leadership teams, executive education cohorts and university programs. No coding experience required. Caps at 15 for hands-on delivery.
Keynote or briefing · 40 min + Q&A
When AI governance becomes a series of approvals, people find ways around it. Karina tells the story of building one that did the opposite: generative AI spreading through a regulated business with no function owning the enterprise risk, three options put to senior leadership, and the decision to widen governance so the affected functions became the decision makers. A cross-functional task force wrote the responsible AI policy with no incremental headcount. It created the governance foundation for the enterprise Microsoft Copilot deployment and subsequent AI use cases.
What they leave with: the council operating model, intake and decision rights, and lifecycle guardrails. Plus the reframe that carried it: from "how do we control AI" to "how do we enable responsible use while protecting customer information, the company and its strategic options."
For CIOs, CTOs, risk, legal, compliance and quality leaders, particularly in regulated industries.
How leaders distinguish increased output from improved outcomes, identify hidden rework, and decide where AI should advise or act. Available as a 45-minute keynote.
Generative AI was spreading through a regulated business, but no single function owned the enterprise risk. I brought three options to senior leadership. We chose to broaden ownership so the functions affected by AI decisions helped make them.
A cross-functional task force developed the responsible AI policy with no incremental headcount. The CEO approved it, and it became the governance foundation for the enterprise Copilot deployment and subsequent AI use cases.
The lesson I share from the stage: governance is an accelerator - it can help an organization move forward when people understand what they can do, who decides, and when to ask for help.
My path has taken me through mathematics, economics, product management, and enterprise technology leadership. Across those disciplines, my focus has been on connecting technology decisions to how organizations work and what they want to achieve.
After eight years leading an AI and digital function in regulated life sciences, I wanted to explore firsthand what changes when AI starts carrying out work. I spent eight weeks building the agent system I now use in my own work.
Building brought leadership questions into sharp focus: What should an agent be allowed to do? How do we assess its work? When must it stop and hand back to a person? Those questions now shape my advisory work, talks, and workshops.
Talk, panel or fireside · 30 to 45 min
AI can help organizations do existing work faster. It can also make possible work they could not previously do: serve an overlooked audience, offer a new service, or pursue an idea that once cost too much to explore. Karina examines how leaders distinguish efficiency gains from new opportunities, question the assumptions behind today’s workflows, and decide where to experiment.
What they leave with: a practical lens for distinguishing efficiency and opportunity use cases, questions to uncover possibilities beyond the current workflow, and a way to test whether an idea creates value worth pursuing.
A strong fit for leadership teams, innovation audiences, and moderated conversations about AI strategy and growth.
Talk, panel or fireside · 30 to 45 min
As AI expands what teams can produce, leaders must become more deliberate about what deserves to be done. Karina explores how work changes when people and agents share responsibilities.
What they leave with: questions to ask when redesigning organizational ways of working. Top leadership capabilities that take your leadership to meet the moment.
One of her most adaptable session, and the strongest fit for panels and moderated conversation.
Talk, panel or fireside · 30 to 45 min
When AI starts carrying out work, leaders must decide how to divide responsibilities, grant authority, assess quality, and handle exceptions. Karina explores how work changes when people and agents share a workflow, and what must be in place before an agent is trusted to act.
What they leave with: a test for which work belongs to an assistant and which belongs to an agent, and the conditions that have to hold before an agent is allowed to act.
Her most adaptable session, and the strongest fit for panels and moderated conversation.
A hands-on AI agent workshop for leaders
Workshop · Half day, or 4 × 90 min virtual
Leaders build a bounded agent prototype during the session, then test it against what decides whether an agent can be trusted in a real workflow: explicit permissions, quality criteria, and escalation rules for when it must stop and hand back to a person. Karina spent eight weeks building the agent system that now runs her own work, and the workshop distills what that taught her about the components of one and its organizational impact.
What they leave with: a working prototype they built themselves, the permission, quality and escalation rules they tested it against, and a clear-eyed list of what their own organization is not yet ready to let an agent do.
For leadership teams, executive education cohorts and university programs. No coding experience required. Caps at 15 for hands-on delivery.
Keynote or briefing · 40 min + Q&A
When AI governance becomes a series of approvals, people find ways around it. Karina tells the story of building one that did the opposite: generative AI spreading through a regulated business with no function owning the enterprise risk, three options put to senior leadership, and the decision to widen governance so the affected functions became the decision makers. A cross-functional task force wrote the responsible AI policy with no incremental headcount. It created the governance foundation for the enterprise Microsoft Copilot deployment and subsequent AI use cases.
What they leave with: the council operating model, intake and decision rights, and lifecycle guardrails. Plus the reframe that carried it: from "how do we control AI" to "how do we enable responsible use while protecting customer information, the company and its strategic options."
For CIOs, CTOs, risk, legal, compliance and quality leaders, particularly in regulated industries.
How leaders distinguish increased output from improved outcomes, identify hidden rework, and decide where AI should advise or act. Available as a 45-minute keynote.