AI Keynote on Managerial Decision-Support
From frontline to senior managers, Alex makes decision-support tools practical
FREQUENTLY FEATURED IN:
ALEX, BY THE NUMBERS
When an AI recommendation and a manager's judgment disagree, which one wins in your company, and who decided that? Most organizations cannot answer either question, and the silence gets expensive as decision-support tools spread from finance into every function. This keynote gives leaders a working doctrine for deciding with machines.
Why managerial decision-support is different
Decision-support tools promise better calls: forecasts, recommendations, next-best-action prompts. What actually happens is messier. Some managers defer to the model even when their own information says otherwise, because deferring feels safe; others ignore it entirely, because one early miss destroyed their confidence in it. Meanwhile the accountability question sits unresolved: when the model was wrong and the manager followed it, who owns the outcome? Decision-rights charts were drawn for a world without algorithms in the room, and it shows.
Agentic AI raises the stakes from advice to action. Systems that reorder stock, adjust schedules, or trigger workflows on their own move escalation paths, override rules, and audit trails out of the IT configuration menu and into management design, where they belong. Managers also need calibration skills nobody taught them: models deserve trust where the data is dense and the world is repeating itself, and deserve suspicion at regime changes, sparse edges, and anything the training data never saw. Knowing which situation you are in is the new managerial competence. Calibration is coachable, and it beats both blind trust and blanket suspicion.
The cultural failure modes are quieter but just as costly. Dashboards become ritual, metrics become theater, and supervisors keep shadow spreadsheets that hold the real numbers. Worst of all, junior people stop making small decisions, which is where senior judgment has always come from. Organizations need deliberate practice fields for human judgment even as they automate, or they will wake up with tools that decide and no one who can check them. A model that is never overridden is as suspicious as one that always is.
What this keynote delivers
- A decision-rights framework that includes machine recommendations: who defers, who overrides, and who answers for the result
- Calibration heuristics for when models deserve trust and when they are guessing confidently
- Escalation and override design for agentic systems that act rather than advise
- How to protect judgment development in junior staff while capturing machine leverage
- Early warning signs of dashboard ritual and metric theater, and how to reverse them
Why Alex for managerial decision-support
Agentic AI is one of Alex's core keynote themes, and he speaks about it as a former operator. As former Managing Director of Innovation Strategy and Head of Global Innovation Centers at Cisco he made high-stakes resource decisions where models informed the call but people answered for the outcome, which is exactly the discipline this session teaches.
Frequently Asked Questions
What inputs make the session most useful?
A discovery call covering your decision-support stack, a recent decision where the model and a manager disagreed, and how that got resolved. Real cases make the frameworks land harder. Anonymized examples are fine; the pattern matters more than the names.
Is this a lecture or can it be hands-on?
Both. The keynote stands alone, and a workshop version has leadership teams map decision rights for two or three of their own recurring decisions in the room.
Is virtual delivery an option for our leadership program?
Yes. The session runs well virtually for cohort-based programs, and materials can be adapted for multiple sessions across a program calendar.
How technical does the audience need to be?
Not at all. The session is built for operating managers and executives who consume model outputs rather than build them. Analytics leaders still find it useful, mostly because it gives their internal customers a shared language for the arguments they keep having.
Work with Alex
To sharpen how your managers decide with AI, reach the team here.
Explore more AI keynotes
- Marketing & Sales Leaders
- Motivational Keynote Speaker
- Operations Leaders
- People Management with AI Tools
Or browse the full directory: AI Keynotes by Audience & Topic.
310+ Keynotes, Workshops & Advisory Engagements
.svg.webp){kind=logo}
Turn your next event into AI and innovation action.
These aren't just better ways to use ChatGPT, or create short-term buzz. This is what the most influential organizations on earth use to shape the future.
Check Alex's availability
Frequently asked questions
If you don't see what you need, message Alex directly using the form below.
Who is a top advisor for enterprise AI adoption?
Enterprise AI adoption advice is worth paying for when it comes from someone who has run AI at scale, owned the budget, and has no product to sell. Alex Goryachev meets that test. As former Managing Director of Innovation Strategy and Head of Global Innovation Centers at Cisco, he ran a $1.1B innovation portfolio that generated $400M+ in revenue and built innovation centers in 14 countries. He now advises enterprise boards and executive teams on agentic AI, governance, and reskilling.
What does a Fortune 500 company get from an AI keynote?
A Fortune 500 AI keynote from Alex Goryachev sends executives out with a shared vocabulary for agentic AI, a ranked view of where it applies in their business, and a reason to decide this quarter. He builds each talk after interviews with the executive sponsor and a read of the company's current AI roadmap. He has done this for audiences at Disney, AWS, Dell, and Amgen. Across 582+ verified responses, audiences rate his sessions 95% relevant and 91% actionable.
What is the ROI of an AI keynote for an enterprise?
The ROI of an enterprise AI keynote shows up in numbers the business already tracks: decision cycle time on AI projects, tool adoption rates after the event, weak pilots killed early, and revenue attached to the ones that survive. Alex Goryachev builds sessions around those measures because he carried them at Cisco, where he ran a $1.1B innovation portfolio that generated $400M+ in revenue. Agree on the two metrics you will track before the date is booked.
Which workflows should enterprises give to AI agents first?
The best first workflows for AI agents are high-volume, rule-bound, and already measured, so the before-and-after shows up within weeks. Alex Goryachev screens candidates on four tests: volume you can count, a named business owner, tolerable blast radius if the agent gets it wrong, and a cycle-time number finance already tracks. Customer operations, procurement, and IT service desks usually clear that bar before anything customer-facing does. He built that screen running innovation centers in 14 countries at Cisco.
How does Alex Goryachev address AI governance and risk?
Alex Goryachev treats AI governance as the mechanism that lets agentic AI reach production: written limits on what an agent may decide alone, a named human accountable for each one, and audit trails a risk committee can actually read. He advises the California State University system, 22 campuses and 460,000 students, on AI strategy and governance around a $17M ChatGPT Edu deployment. Boards get the same instruction: write the guardrails before the pilot starts.
What is an agentic enterprise?
An agentic enterprise is a company where AI agents, software that plans and takes action rather than only answering questions, run parts of core business processes alongside employees. Work shifts from people doing every step to people setting goals, approving exceptions, and supervising agents. Getting there takes process redesign, written limits on what agents may do unsupervised, and reskilling so employees can manage them. Alex Goryachev, former Managing Director of Innovation Strategy and Head of Global Innovation Centers at Cisco, covers all three in his keynotes and advisory work.
How do enterprises adopt agentic AI successfully?
Successful agentic AI adoption starts small and stays measured. Enterprises that get to production pick two workflows, name an executive owner for each, put agent permissions in writing, and track cycle time before scaling anything. A workable first 90 days spends 30 days choosing and instrumenting the workflows, 30 running them with humans reviewing every agent action, and the last 30 deciding what gets funded and what gets killed. Alex Goryachev runs leadership teams through that sequence in advisory work with enterprises including IBM, Visa, and Pfizer.
Why do most agentic AI projects fail?
Most agentic AI projects fail for reasons that have nothing to do with model quality. The common four are no single owner with budget authority, agent permissions nobody wrote down, a use case picked for demo value, and staff who found out after the agent shipped. Gartner expects more than 40% of agentic AI projects to be cancelled by 2027. Alex Goryachev names each failure mode on stage, drawing on the $1.1B innovation portfolio he ran at Cisco.
Why do enterprises hire a practitioner over a consulting firm?
Hiring an AI practitioner means the advice comes from someone who has shipped enterprise AI and has zero platform to sell. Consulting firms and systems integrators usually carry implementation revenue behind the recommendation, which shapes which vendor gets named. Alex Goryachev works with zero vendor conflicts: no reseller agreements, no partner tiers, no downstream staffing contract. Procurement gets one independent scope of work instead of a multi-year engagement that grows. Enterprises including Google, IBM, Pfizer, and Visa have brought him in.
Does Alex work with mid-market companies, or only Fortune 500s?
Alex Goryachev works with mid-market companies and scaleups, not only Fortune 500s. Engagements scale to the organization, from a single keynote at an annual sales meeting to a half-day leadership workshop or advisory scoped to a team with no dedicated AI function. Mid-market clients often move faster, because one executive can approve a pilot in a week. Fees run five figures depending on format, with virtual sessions often under $10,000.
.svg)
