AI in Performance Reviews: A Keynote on Judgment, Fairness, and the Manager's Job
From appraisal systems to AI tools, Alex makes reviews smarter and more transparent
FREQUENTLY FEATURED IN:
ALEX, BY THE NUMBERS
Who actually wrote your last performance review, the manager or the model? AI is quietly drafting evaluations, summarizing feedback, and scoring goals inside companies that never made a deliberate decision to allow it. This keynote makes that decision deliberate.
Why performance reviews are different
Reviews shape pay, promotion, and dignity, which makes them the most trust-sensitive documents a company produces. AI-drafted reviews read polished and slightly generic, and employees can tell. The moment someone suspects a machine wrote their evaluation, the message received is that they were not worth an hour of their manager's attention. The efficiency gain is real; so is the trust cost, and only one of them shows up in the HR dashboard.
Fairness cuts in both directions. Used well, AI can reduce recency bias and smooth wild inconsistency between managers. Used carelessly, it launders existing bias into neutral-sounding prose. The data feeding these tools often measures visibility rather than value, which quietly penalizes remote employees and quiet contributors whose work leaves fewer digital traces.
There is also a record-keeping dimension. Evaluations become evidence in employment disputes, and questions about what role automated tools played in a rating are coming. HR needs a defensible position now: where AI may assist, where managers must own every word, and what employees are told about the process. Calibration is where all of this converges. Rating meetings already suffer from anchoring and horse-trading, and machine-generated summaries add a new distortion: whoever controls the prompt controls the narrative that arrives in the room. HR needs visibility into how ratings were assembled, not just what they concluded. The safest posture treats AI output as one input among several, with the manager accountable for the final words, and a periodic audit comparing outcomes across teams and tools before patterns harden into precedent.
What this keynote delivers
- A clear line between AI as drafting aid and AI as decision-maker, and why crossing it changes everything
- The bias question asked properly: where automation reduces inconsistency and where it hides it
- Disclosure choices, including what to tell employees about AI's role in their reviews
- Manager craft for keeping specificity and judgment when drafting becomes free
- Governance guardrails HR can set this quarter without banning useful tools
Why Alex for performance reviews
AI governance is one of Alex's core themes, and he serves on the AI Working Group advising the California State University system on responsible AI adoption. He is independent, with no stake in any HR software vendor, which lets him talk about these tools plainly. That governance grounding means the session goes past tool tips into the decisions HR leaders actually face: policy, disclosure, and where accountability must stay human.
Frequently Asked Questions
Our review practices are sensitive. Will they stay in the room?
Yes. NDAs are routine for sessions of this kind, and nothing your team shares is repeated or reused in other engagements.
Is Alex affiliated with any performance management vendor?
No. He sells nothing from the stage and holds no vendor relationships, so his read on the tooling market is unclouded by commissions.
Who should attend, HR only or all people managers?
The session works best with both. HR sets policy, but managers write the reviews, and putting them in the same room surfaces the gaps between policy and practice quickly.
What follow-up does our HR team receive?
A recap of the frameworks and the governance questions raised, formatted so your HR leadership can turn it into policy discussions without re-teaching the session. Many HR teams use it to brief their people-leader community in the weeks after, so the standards discussed on stage become shared practice rather than one team's notes.
Work with Alex
To put judgment back at the center of your review cycle, ask about availability.
Explore more AI keynotes
- AI in Recruiting
- AI Literacy for All
- AI-Powered Innovation Culture
- AI-Powered Learning & Development
- AI-Powered Team Leadership
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)
