ISTM 629 Cumulative Final Study Guide

Digital Platforms for HR — Units 1 through 6 — Multiple Choice, Matching, and True/False

At a Glance
• Built from all course materials: every article and case notes doc, every lecture and video
notes doc, and the actual quiz questions from Modules 1 through 4, so the guide is tuned
to how this professor actually asks.
• Coverage: Unit 1 (digital world and emerging tech), Unit 2 (data, analytics, AI), Unit 3
(HRIS systems and cybersecurity), Unit 4 (agile HR, AI ethics, and the law), Unit 5
(BPM and RPA), Unit 6 (future of HR, Beamery, and IT project management).
• Each unit gives you key terms with tight definitions, the named lists and frameworks
(prime matching material), the numbers worth knowing cold, and the case facts the
quizzes have pulled from.
• Then three exam-ready sections: how this professor writes questions, a master matching
bank, and a true/false trap list built from real quiz wording.
• Night before: read the question-pattern section, the matching bank, the trap list, and the
Fast Recall box. That is your final pass.

How This Professor Writes Questions (Learned from Your Real Quizzes)
• The “All of these” pattern. When an article names several areas, the quiz lists two or three
of them plus “All of these,” and “All of these” is the answer. Seen with: agile change areas
(appraisals, coaching, L&D), data-driven HR decisions (pay, performance, ATS), examples
of AI (computer vision, NLP, ML, pattern recognition), and the change model elements
(processes, structure, skills). If every listed option appears in the source's list, take “All of
these.”
• Exact acronym definitions. ERP, GIGO, SaaS, RPA, HRIS, WINS, DELTA, TIM WOOD
U, DMAIC. The wrong options are plausible-sounding fakes (“Enterprise record of personnel
systems”). Know the real expansions letter by letter.
• One-word T/F hinges. True/false items flip a single word in a definition: system of record
vs. system of engagement, structured vs. unstructured, ERP vs. best-of-breed, formatted vs.
unformatted packet. Read for the swapped word.
• Article attribution matters. Questions open with “According to the article X…” and the
distractors are true statements from a different reading. Keep each claim attached to its
source (WINS belongs to Baier, not Chamorro-Premuzic; five considerations belong to
Chamorro-Premuzic, not Baier).
• Multiple Answer with one or two poison options. Select-all questions include correct
options phrased positively and one or two poison options that maximize efficiency by
removing humans (“eliminate all human interaction,” “focus primarily on replacing
employees,” “assume employees will become unnecessary”). Every reading in this course
rejects the replace-the-human framing. Never select the dehumanizing option.
• NOT questions swap categories. “Which is NOT a challenge” style questions plant an item
from a neighboring list in the same reading. Real example: “more renting, less buying” is one
of SHRM's five TRENDS, planted among the five CHALLENGES. Keep parallel lists from
the same source separate.
• Precise case details. Quizzes have pulled named people and specific decisions from cases:
Jim Roth putting agents under contact center managers, Brian Millham hiring AI salespeople
during layoffs, Target's pregnancy insight being predictive analytics. Case sections below
flag these.

Unit 1: Our Digital World and Emerging Technologies

Key Terms and Definitions
ERP Enterprise Resource Planning system. Integrates all business
functions into one unified corporate database everyone reads
from and writes to; extensible; enables horizontal end-to-end
processes. Quiz fact: ERPs CAN be used to foster social
responsibility.
Digital ecosystem “An interdependent group of actors (enterprises, people,
things) sharing standardized digital platforms to achieve a
mutually beneficial purpose.” That exact sentence was a True
on the quiz. LEGO is the course's digital ecosystem example.
Gartner Hype Cycle Five stages in order: Innovation Trigger, Peak of Inflated
Expectations, Trough of Disillusionment, Slope of
Enlightenment, Plateau of Productivity. Quiz-keyed purpose: to
understand how technology will impact certain industries.
VOIP is the took-years-to-mature example; in the 2018 chart
nothing sat on the plateau. The professor places current AI at
the peak of inflated expectations.
Shape-shifting Bob Johansen's future org: MORE distributed authority, LESS
organization (Johansen) centralized (that direction was the quiz answer). By 2030
organizations are liquid: no fixed center, hierarchies come and
go, positions in constant flux.
System of record Backend. Structured, stable, reliable, transactional, single
source of truth (accounting, core HR). Where enterprise data is
stored, used by internal employees.
System of engagement Frontend. Dynamic, conversational, quick-adapting, customer-
facing. T/F trap: “systems of engagement are where data from
enterprise systems are stored, used by internal employees” is
FALSE; that sentence describes a system of record.
SaaS / PaaS / IaaS Cloud service pyramid. SaaS: most used, complete hosted
 software ([Link], Waze) — “[Link] is SaaS”
and “SaaS is software hosted in the cloud” are both True. PaaS:
for developers. IaaS: raw computing. Responsibility shrinks as
you go up: traditional IT (you manage everything) → IaaS →
PaaS → SaaS (vendor manages everything).
Cloud types Public (accessible), private (secure), hybrid (mixed).
Virtualization Replacing hardware with software — that phrase is the quiz
answer. Types: server, network, desktop, application, storage.
Hypervisor runs it; consolidation puts about ten servers on one.
Moore's Law Processing power doubles every 18 to 24 months. Companion
fact: an iPhone has roughly 100,000 times the processing
power of the Apollo moon computer.
IT vs. IS Information Technology = hardware and software. Information
Systems = IT plus people, data, processes, and objectives. Four
IT categories: hardware, software,
networking/telecommunications, databases.
Packet A FORMATTED unit of data carried by a packet-switched
network (quiz answer; “unformatted” is the trap). A large page
is broken into thousands of packets that take multiple
independent paths and reassemble at the destination in about a
second (True on quiz).
IP address vs. URL [Link] is a URL, NOT an IP address (quiz
False). DNS translates URLs into IP addresses (e.g.,
[Link]).
Client-server / networks The internet is the world's largest client-server network (quiz
True). LAN = building network; WAN = network of networks;
the internet is the largest WAN.
Fourth Industrial Brings together digital, physical, and biological systems; its
Revolution signature is that it changes US, not just what we do. Lineage:
steam → electricity/assembly line →
computers/communications → the fourth.
Superagents (Bersin) AI systems that own an entire end-to-end process, like
onboarding or talent acquisition — that full phrasing was the
quiz answer. Progression: assistants → agents → Superagents.
NOT individual productivity tools and NOT young employees.
Service as software Salesforce's phrase for the business-model shift where software
stops assisting labor and becomes the labor; licensing seats
gives way to consumption pricing.
Agents vs. copilots Copilots assist and suggest; agents act autonomously within
 user-defined guardrails, retrieve data on demand, and hand off
to humans with a summary.

Bersin, The Great Reinvention of HR (Jan 2026)

• HR's dual identity: strategic value creator vs. administrative “company police.” The AI
reinvention resolves it toward full-stack HR.
• 30 to 40 percent of existing HR jobs are automatable with low effort (per Galileo, his AI
agent with task data on 250+ HR jobs) — yet the profession will NOT shrink: HR postings
rose 60 percent in five years; the mix changes and wages likely rise.
• Employee-to-HR ratio shifts from about 100:1 toward 200, 300, even 400:1.
• Car analogy: stop making the driver more productive; optimize the outcome (passenger
mobility). The driver goes away, safety jumps, new business models appear (pay by the
mile).
• Chipotle is his food service proof: AI-driven hiring and scheduling speed directly raises
revenue.
• New risk surface: data quality, explainability, trust. One automated error hits thousands
instantly. HR now owns AI bias — evidenced by the two vendor lawsuits: Workday and
Eightfold.
Salesforce Agentforce Case (Quiz Favorite)
• Jim Roth (President of Salesforce Customer Success) decided AI agents are managed by the
SAME contact center managers who oversee human reps — not IT, not a new digital labor
division, not an external vendor. The business owns agent quality.
• Brian Millham (COO): hiring several thousand new salespeople focused on selling AI
products while planning early-2025 layoffs = workforce transformation involves job
displacement AND creation of new, differently-skilled roles simultaneously (quiz answer).
• Customer zero results: 380,000 conversations, 84 percent resolution, 2 percent escalation;
human escalations halved in week one. About $2 per agent conversation vs. roughly $8 to $9
per human conversation.
• Trust: number one value since the 1999 founding; Einstein Trust Layer = data masking, zero
data retention, audit trails; no customer data goes into an LLM.
• Marc Benioff at Davos: these are “the last CEOs who will manage only humans.” Proof
points: Wiley ($230,000 saved, 40 percent improvement) and OpenTable (shorter calls, more
meaningful engagement). Implementations about 16 times faster (live in four to six weeks);
$6 trillion labor market target; goal of one billion agents.
Infrastructure Videos
• A Packet's Tale: England request → copper wires → Telehouse North (London, England's
main hub) → undersea fiber as light → 60 Hudson Street NYC (largest east coast hub) →
LA server; return via One Wilshire (LA). Packets don't care how they get there, only that
they arrive fast; whole trip about one second.
• Inside a Google Data Center: Joe Cava, VP of Data Centers, South Carolina site (one node
in a global network), “brains and engine of the internet.” Layered security rising toward the
center: campus badge list → building → secure corridor (badge PLUS biometric iris scan) →
data center floor (highest; under-floor laser intrusion detection). Floor runs warm at about
80°F for efficiency. Failed drives: end-to-end chain of custody; unverified drives are crushed,
then shredded in an industrial wood chipper. Hot-aisle containment cooling with water-filled
copper coils; cooling tech changed five times in about 6.5 years.
HR Technology: A Complete Guide
• Three drivers of HR tech evolution: cloud computing, machine learning, AI.
• Scope ladder: HRIS (admin) → HRMS (adds recruiting, performance, reporting) → HCM
(adds succession, comp planning, analytics, talent development). Note the Unit 3 lecture
version differs slightly (HRIS → HCM → HRMS with payroll/time); if the question cites the
lecture, HRMS is the payroll/time tier.
• Core HR = payroll, benefits administration, employee records, compliance reporting. Top
challenges: system integration, global compliance, proving ROI, cost, user adoption, PII
security.

Unit 2: Data, Analytics, and AI

Key Terms and Definitions
GIGO Garbage In, Garbage Out (quiz answer verbatim). Fix data at
the source; line-of-business staff now do cleanup, not just IT.
Data quality dimensions: Complete, Accurate, Available,
Timely.
Four Vs of Big Data Volume, Velocity, Variety, Veracity. Quiz phrasing: Big Data
is “best defined as” the four Vs — NOT “vast amounts of
information on the internet.” Processed with Hadoop and
Apache Spark.
Structured vs. Structured fits tables (relational databases, primary/foreign
unstructured keys). Unstructured = text, social posts, images, audio —
handled by NoSQL (column, document, graph, key/value).
Quiz False: “a relational database is primarily used to manage
unstructured data.”
Operational DB vs. data Operational/transaction systems hold live data and stay fast; the
warehouse data warehouse holds copied data for analysis and what-if
questions; fed by ETL (Extract, Transform, Load); data marts
are department slices (e.g., an HR mart with purchased
demographic and salary data).
Data lake Raw structured plus unstructured data, loosely coupled, stored
without a predetermined use case. The lake's real challenge is
 consistency and completeness standards, not just quality.
Gartner's four analytics Descriptive (what happened) → Diagnostic (why) →
types Predictive (what will happen) → Prescriptive (how to make it
happen). Target's pregnancy insight = PREDICTIVE analytics
(quiz answer).
Data mining Discovering patterns and correlations in data (Target
pregnancy predictor). Prerequisite: clean, consistent data. Mine
with an orientation — a road trip needs a NAV — but keep an
experimental lane open.
AI / ML / DL nesting AI contains Machine Learning, which contains Deep Learning.
Cognitive computing = simulating human thought (IBM
Watson). Supervised learning = training on labeled data
(millions of labeled cat photos).
HIPPO Highest Paid Person's Opinion — gut-based decisions the
evidence-based movement replaces. Roundtable nuance:
decisions can still be roughly 20 percent analytics, 80 percent
experience; data supplements judgment.
Data scientist Rare, highly paid, spends up to 70 percent of time cleaning
data; profiled as a storyteller who communicates at every level,
not just a statistician.

What Is AI, Exactly (Video)

• Term coined by John McCarthy, 1955, Dartmouth Conference (Turing's work predates the
defined field). Definition: any aspect of learning or intelligence can be described precisely
enough for a machine to simulate it — and improve itself.
• Examples of AI: machine learning, computer vision, natural language processing, robotics,
pattern recognition, knowledge management — quiz answer was All of these.
• Strong AI (simulates the brain, gives insight into it — nowhere near); Weak AI (behaves
human-like, no brain insight — IBM Deep Blue); middle ground (inspired by human
reasoning, not bound to it — IBM Watson with percent-confidence answers). Neural
networks are a subset of machine learning. Most AI = expert systems (human knowledge in a
computer); AlphaGo = first proven non-expert system (DeepMind, Demis Hassabis).
What Is Big Data (Intel Video)
• Dawn of time to ~2003: about 5 exabytes; 2012: 2.7 zettabytes (500 times all data before
2003). Growth driver: continuous generation from more sources and devices, much of it
unstructured; some arrives too fast to store first.
• Gartner: only 10 to 15 percent of organizations will fully exploit big data, but those that do
outperform unprepared competitors by 20 percent across major financial metrics.
• Truck-dispatch example: the closest truck is not necessarily the best truck (traffic jam, no
cargo space, steep grades) — optimal decisions need real-time analysis of multiple sources
(GPS, traffic, RFID).
Competing on Talent Analytics (Davenport, Harris, Shapiro, HBR 2010)
• Thesis: favor analytics over gut instinct — people are a company's greatest asset and largest
expense — but no company studied went analytics-only.
• Six uses, simple to sophisticated: (1) Human-Capital Facts — JetBlue's invented
crewmember net promoter score; (2) Analytical HR — Lockheed Martin; (3) Human-Capital
Investment Analysis — Sysco, retention 65 to 85 percent, about $50M saved; (4) Workforce
Forecasts — Dow Chemical what-if modeling; (5) Talent Value Model — Google
intervening on the lowest 5 percent; (6) Talent Supply Chain — real-time staffing decisions.
• DELTA: Data (sufficient, not perfect), Enterprise orientation (break HR silos), Leadership
(single most important factor), Targets, Analysts.
• Signature numbers: a 0.1 percent engagement gain = $100,000 in store annual income at Best
Buy; Google cut interviews from ten to five; Cognizant bloggers 10 percent higher
performance; AC Milan: 60,000 data points per player. Project Oxygen: Google built the
good-manager definition from scratch — eight behaviors plus five to avoid.
• Named mistakes: treating people like interchangeable widgets, zombie metrics, gameable
few-metric systems, demanding 100 percent accuracy, judging on grades and test scores
alone.
Apturja Power Limited (Ivey Case)
• Cast: Kaushik (business head) → Chaubey (HR head, 21 years) → Sharma (recent MBA
who took an HR analytics course) does the analysis.
• Problem: first-year attrition among ~1,000 annual fresh graduate hires rising — resignations
455 (FY18) to 545 (FY19); about half of 2,000 two-year hires left within a year — despite
above-market pay, five-star orientation, townships.
• Exhibit 4 paradox: performance rate up (764→778) and training efficiency up (808→812)
while retention fell. The star performer resigned citing autonomy, burdensome targets, and
monotony — the driver is job design and meaning, not pay. Chaubey chooses data over
instinct.
• Five recruitment channels at roughly 20 percent each: campus (~$1,000/visit), online portals,
APL Ward Program ($20/hire), agency/union (under $150/yr), LinkedIn (~$200/cycle).
Roundtable (McQuaid and Sutton)
• Manufacturing-to-service shift: about 70 percent of GDP is service, where input and output
both vary — you can count a riveter's rivets but cannot measure a barista that way.
Accountant (backward-looking ranking) vs. motivator (building assessment systems that
improve performance).
• Trust paradox: management always reports trust exists; where it is absent people still report
it. Pair with Everybody Lies: watch behavior, not surveys. Frame first — Einstein: 55 of 60
minutes on the question. Storage ladder: relational DB → data mart → data warehouse →
 data lake. Closing charge: who are our people, what are their competencies, and what data
and systems would answer that.

Unit 3: HRIS Systems, Vendors, and Cybersecurity

Key Terms and Definitions
Integrated suite (ERP) vs. Integrated ERP-style suite (SAP, Oracle, Workday, UKG): one
best of breed source of truth, convenient, but costly and easily underused.
Best of breed: the best specialist application per area — strong
at single tasks, painful to integrate. Quiz False: “ERPs like
SAP or PeopleSoft are considered best-of-breed applications.”
They are the opposite. No single correct approach exists; it
depends on the organization.
Top buyer requirement Cloud-based — the number one requirement of HR software
buyers (quiz answer), followed by payroll and benefits
administration.
Scott Morton Org Change Five linked elements: strategies, business processes, structure,
Model (1991) individual skills and roles, technology — all connected with
double-headed arrows. Change one and the others must move.
Quiz pattern: “when implementing new technology,
consider…” = All of these. Sequencing rule: optimize the
process first or you pave the cow path.
Don't pave the cow path Automating a bad process just speeds up the mess; fix the
process before laying technology over it. Viggiano's version:
“you should not Web-enable bad processes” (also a Novartis
lesson).
Ransomware It ENCRYPTS files on the hard drive (quiz answer). Stages:
spam attachment → malware executes → files encrypted →
computer locked → ransom demand (bitcoin).
Attack and malware types Malware: viruses, Trojan horses, rootkits, back doors. Attacks:
phishing, spear phishing, social engineering, malvertising,
ransomware, DDoS. DDoS = non-intrusive (overwhelms, no
theft); intrusion = steals or damages. Quiz True: any smart
wireless-enabled appliance is a security threat.
Hacker hats Black (malicious), white (defensive, often reformed), gray
(mostly good, questionable methods). Quiz False: “all hackers
are unethical.”
Defenses Strong passwords on all networked devices, multifactor
authentication (two or more factors), access control lists (read,
write, execute per object).
 Waterfall vs. agile (SDLC) Waterfall: seven linear steps — planning, analysis, design,
development, testing, implementation, maintenance. Agile:
iterative sprints of two to four weeks (30-day typical), 24-hour
daily scrum, user stories in a backlog, scrum masters instead of
project managers, output is a potentially shippable product
increment.
User resistance tactics Use only necessary features; continue replaced work processes;
(four) refuse to learn beyond basics; blame every error on the new
technology. Unmanaged, the investment sticks in the lethargy
stage.
Web layers Surface web (public, ~50 billion pages indexed); deep web
(access-controlled, not indexed, roughly 500 times larger); dark
web (needs Tor — The Onion Router, built by the U.S. Naval
Research Laboratory, launched 2003).

Gartner Magic Quadrant for Cloud HCM Suites (Oct 2024)

• Quiz-corrected fact: the Magic Quadrant DOES present the strengths and weaknesses
(strengths and cautions) of individual technology vendors — True. You missed this one on
the quiz; do not miss it twice.
• Placements: Leaders — Dayforce, Oracle, SAP, UKG, Workday. Challengers — ADP,
Darwinbox. Niche Players — Cegid, Cornerstone, Infor, Yonyou. Visionaries — NONE this
edition (weight shifted to execution of administrative HR).
• Axes: ability to execute and completeness of vision. Market: grew 14.4 percent in 2023 to
$26.3 billion; top five Leaders hold about 62.1 percent of revenue.
• 2027 assumptions: 75 percent of midsize/large enterprises on a cloud HCM suite (still
needing other tools for 20 to 30 percent of needs); employees initiate 30 percent of
transactions by natural language command.
• Vendor hooks: SAP = geographic king (core HR in 104 countries, 51 payroll localizations,
Joule assistant). Workday = best customer experience (96 percent would recommend),
payroll in only five countries. UKG = highest WFM and IHRSM (Bryte AI). Oracle =
highest innovation, prehire strength. Dayforce = WFM/payroll strength, rebranded from
Ceridian 2024. ADP = highest-rated payroll, US/Canada only.
SHRM: Leveraging HR Technology for Competitive Advantage (Johnson and
Gueutal)
• Five keys: (1) use new technology as an opportunity to change HR (quiz True: implementing
a new system is a chance to reexamine how a process is done; also True that new tech can
change HR work processes); (2) make employees self-reliant; (3) communicate when
monitoring performance; (4) use more data to get more answers; (5) mind appearance and
functionality — for many employees the HR web site IS HR.
• Three adoption strategies: single integrated vendor (PeopleSoft, Lawson named), best of
breed, or outsource the infrastructure (ADP named).
• Five challenges: new HR skills and roles; increasing expectations and demand for data; HR
distance and isolation (less face-to-face); faster may not mean better; adverse impact (lower
socio-economic groups have less computer access). Quiz NOT-question answer: “more
renting, less buying of services” is NOT a challenge — it is one of the five TRENDS (social
networking, compliance expansion, renting over buying/SaaS, BI and dashboards,
transparency vs. privacy).
• Research sidebars: e-recruiting increases applicant VOLUME but not QUALITY; e-learning
works best with peer interaction.
Novartis HRIS (Stanford Case)
• Novartis formed December 1996 from the Ciba-Geigy and Sandoz merger ($22B, largest of
its time); name means “new skills”; goal: premier talent machine by 2005; CEO Dan Vasella.
• The ECN approved CHF 78 million for a global HRIS in September 2002. Core point: far
more than an IT project — the enabler of HR's transformation from transaction processor to
strategic partner.
• Legacy mess: PeopleSoft, SAP, local systems, or none; no consolidated database; managers
could not even say how many HR associates worked at Novartis. Chosen platform: SAP R/3
(transactions) plus SAP Business Warehouse (reporting) plus an HR portal.
• Time-shift target: administration 60 percent → 40 → 20; strategy and consulting grow to fill
it. Talent gap: aspiration of two ready-now successors per leadership role and 70 percent
internal fill vs. the 2003 reality of one and 42 percent.
• Lessons: simplify and standardize BEFORE rollout; under-promise and over-deliver; senior
leadership must drive, not just fund; do not Web-enable bad processes; end users must feel
ownership. Closing line: “What is it that we don't know that we don't know?”
Viggiano Interview (Buyer's View)
• “Brochure wear”: vendors sell the big words (chatbots, RPA, AI) but it is three to five years
from mainstream; neither the talent nor the analytics are mature yet.
• The pendulum: ERP → best-of-breed → ERP → cloud — the honest answer is a federated
mix. His reframe: application-centric to data-centric; the real payoff of a suite is the common
data model, which is what actually unlocks AI and the descriptive-to-predictive-to-
prescriptive climb.
• Keyword-scanning ATS = “Neanderthal approach”; the fix is a semantic model / taxonomy
mapping how roles relate — but you must build it yourself for your industry.

Unit 4: Agile HR, Human-AI Collaboration, Ethics, and the Law

HR Goes Agile (Cappelli and Tavis, HBR 2018) — Heavily Quizzed
• Agile is an ITERATIVE process (quiz answer; not deductive, waterfall, or redundant).
Waterfall = linear rather than flexible and adaptive.
• Biggest agile-driven changes: All of these (performance appraisals, coaching, L&D) — the
six change areas are appraisals, coaching, teams, compensation, recruiting, and L&D. Four
 team sub-issues: multidirectional feedback, upward feedback, frontline decision rights,
complex team dynamics.
• Post-WWII planning reasons (multiple answer): rules-based and internally consistent
talent practices; recruited lifers, rotational assignments, groomed years ahead; consistent with
core businesses taking the long view. The poison option: “HR was focused on recruiting
minorities” — not a reason. “The bureaucracy was the point.”
• Upward feedback from employees to team leaders is highly valued in agile organizations
(quiz True). Mitre's finding: the critical factor is managers explicitly saying they want and
appreciate comments.
• AI in L&D: True that AI can target training by matching employees' backgrounds, expected
career trajectory, and training programs completed — that is IBM's exact system.
• Terms: agile lite (principles without all the tech-world tools); scrum (from rugby);
retrospectives (agile after-action reviews); triple-S teams (small, stable, self-managed);
servant leadership (coaching over monitoring); Cisco's “breadcrumbs.”
• Numbers: 79 percent of executives rated agile performance management a high priority
(2017 Deloitte); 90+ percent of IT organizations already use agile; J&J pilot went 20 to 46
percent manager participation with 3,000 pieces of feedback in three months; Regeneron
built four appraisal processes; Intuit started 2009, took four years; VersionOne survey top
scaling challenge = company culture at odds with agile values (63 percent), top success
element = internal agile coaches (52 percent).
• Company hooks: BMO (Lynn Roger: “speed is the new business currency”; embedded agile
coaches); GE (FastWorks; 2015 digital-division agile recruiting with head count manager,
rotating hiring managers, scrum master, kanban board); IBM (Diane Gherson; AI career
advice); Patagonia (eliminated annual raises); Rent the Runway (Jennifer Hyman rolled
bonuses into base pay, “untangling the two”); Macy's (spot bonuses); DigitalOcean (on-site
coach, twice-a-year salary adjustments, only top 1 percent merit pay); P&G (eliminated talent
calibration sessions); Pepsi (quarterly succession updates). Exceptions where agile may not
fit: accountants, nuclear control-room operators, surgeons.
Set Your Team Up to Collaborate with AI Successfully (Chamorro-Premuzic)
• Five considerations: (1) develop your AI augmentation strategy; (2) evaluate OUTPUT
rather than input; (3) build the skills AI is unlikely to master (empathy, EQ, curiosity,
judgment, adaptability); (4) invest in mid-level managers; (5) promote a healthy dose of AI
experimentation (culture where learning from failure is encouraged).
• Quiz-verified correct selections: redirect time toward relationship-building, empathy,
higher-value human work; identify how roles and skills will evolve; use AI to automate
repetitive, predictable tasks; develop curiosity and better questions; recognize human
empathy, judgment, adaptability as competitive advantages; invest in mid-level managers.
Poison options: eliminating all human interaction; assuming employees become unnecessary;
reducing emphasis on empathy.
• Signature line: knowledge workers are less likely to lose their jobs to AI than to another
human using AI. Recruiter anchor: up to 40 percent time savings on keyword scanning,
editing job descriptions, fixing typos. Data points: Microsoft $738B AI market forecast; The
 Economist — 80 percent of US and China organizations use AI daily; ManpowerGroup —
55 percent plan headcount increases because of AI.
Your Organization Isn't Designed to Work with Gen AI (Baier, DeLallo, Sviokla)
• The fundamental flaw: treating gen AI as traditional automation when it is an assistive
agent that gets smarter, and makes humans smarter, over time.
• WINS work = Words, Images, Numbers, and Sounds — processes heavy in WINS work are
ripe for gen AI. This acronym belongs to THIS article only.
• Designing for Dialogue, three components: task analysis (who leads each task, human or
AI), interaction protocols (ground rules — AI flags anomalies for human judgment;
suggestions are starting points, not commands), feedback loops (error reporting, success
metrics, satisfaction surveys, AI explanations).
• Five steps: identify high-WINS processes → task analysis → design protocols and feedback
→ train the team → evaluate, adjust, scale.
• Quiz-verified HR-leader actions: preserve human leadership for emotional intelligence,
nuanced judgment, complex decisions; identify WINS work; continuously monitor AI
performance and employee feedback; train employees to collaborate with AI and build trust.
Poison option: focus primarily on replacing employees for labor savings.
• Jerry case: $450M car ownership company; humans answering 54 percent of inquiries
within 24 hours → AI leading 100 percent, answering 96+ percent within 30 seconds; about
$4M projected annual savings; scaled WITHOUT cutting headcount. BPR context: Michael
Hammer, 1990, “Reengineering Work: Don't Automate, Obliterate” — BPR is top-down
with static task assignment; gen AI needs the dynamic, bidirectional alternative.
The Legal and Ethical Implications of Using AI in Hiring (Dattner et al.)
• Core rule to memorize: it is unlawful to take an adverse action based on protected or private
characteristics NO MATTER HOW they were learned or inferred.
• History anchor: psychometrics in use well over 100 years; spread via the US Military's Army
Alpha. Three traditional categories: cognitive ability/intelligence, personality/temperament,
mental health/clinical.
• Four new tool categories: game-based assessments; social media scraping bots; linguistic
analysis of writing samples; video interviews analyzing speech, tone, emotion, nonverbal
behavior. Old tools were validated and published; new ones emerged as tech innovations
with unclear validity. The article raises three privacy temptation questions (personal
attributes, lifestyle, disabilities) and deliberately does not resolve them.
Workday: Navigating the AI Bias Dilemma (Ivey Case)
• February 2023: Derek Mobley — over 40, Black, with a finance degree — sued Workday
alleging its AI recruitment tools discriminated against Black, older, and disabled applicants;
applications rejected within an hour. The EEOC filed an amicus brief supporting him.
• Three statutes: Civil Rights Act, Age Discrimination in Employment Act, California's Fair
Employment and Housing Act. Two theories: intentional discrimination and disparate
impact. Three bases: race, age, disability.
• Pivotal argument: AI VENDORS should bear legal responsibility if their algorithms
contributed to discriminatory hiring, whether bias was deliberate or unintentional. Why such
suits are rare: case complexity, and candidates not knowing an algorithm evaluated them.
• Workday facts: founded 2005 by ex-PeopleSoft executives Dave Duffield and Aneel Bhusri;
Pleasanton, CA; NYSE: WDAY (2012); by 2024, 16,000+ employees, $6.2B revenue, and
over 70 percent of the Fortune 500 use its tools.
The Ethics Trio and Related Lecture Content
• Cathy O'Neil, Weapons of Math Destruction: algorithms are opinions embedded in code;
bias can live in the algorithm, the historical training data, or both; WMDs are opaque, scaled,
and damaging.
• Shoshana Zuboff, surveillance capitalism: individuals become raw material for data-broker
profiles; behavior is extracted and monetized.
• Eli Pariser, filter bubbles: personalization narrows what you see, now supercharged into
confirmation-bias ecosystems driving polarization.
• Also from lecture: HireVue's facial-scanning backlash; Jobvite 2018 Recruiting Nation
Survey on social sleuthing; AGI = human-level across tasks, ASI = beyond any human; the
professor's through-line — technology far outruns the regulations and the ethics.

Unit 5: Business Process Management and RPA

Key Terms and Definitions
Cycle time Cycle time = process time + delay time. Most slow processes
are dominated by DELAY (waiting), not work. Async = no
waiting for a response (email); sync = real time (phone).
TIM WOOD U The eight wastes: Transport, Inventory, Motion, Waiting,
Over-processing, Over-production, Defects — plus Dwane
Lay's eighth, Underused people. Lay via Deming: approvals are
always waste; people don't fail, systems fail.
Kaizen / Lean / Six Kaizen = continuous “good change” (Toyota). Lean came from
Sigma / DMAIC Toyota; Six Sigma came from Motorola (data-driven defect
reduction). DMAIC = Define, Measure, Analyze, Improve,
Control.
BPMN and flowcharts Flowchart shapes: terminator, process, decision (needs two or
more exits), data, connector. BPMN standardizes them and can
drive a real engine. Gateways: exclusive (either/or) vs. parallel
(the plus sign). Swimlanes show who does what and where
handoffs happen.
As-is / to-be / Gemba Map the as-is before designing the to-be; validate the as-is on
the floor where work actually happens (the Gemba — Ed
 Beverly). Beverly's three must-have HR tools: process
mapping, the seven wastes, root cause analysis (ask why).
Hammer and Champy Organize around outcomes; capture information once at the
BPR source; put decisions where the work is done. Radical redesign
for dramatic improvement.
Don't pave the cow path Fix the process before automating; automating a bad process
just speeds up the mess. UiPath's version: optimize a process
before automating it.
RPA Software robot (not physical) that takes over standard,
repetitive, rule-based tasks. Non-invasive: sits on top of
existing systems, no rip-and-replace. Best HR fits: multi-
system processes (onboarding), rule-based processes (payroll),
recurring data/compliance errors. RPA is NOT real AI (lecture
point).

EY: RPA for HR and Payroll — the Numbers and the Grades
• Headline stats: 65 percent cost savings vs. an offshore FTE; 65 percent of HR rules-based
processes automatable; 93 percent of HR employees' time on repetitive tasks; 20 to 35
percent overall saving potential. Processing time: about 2 hours down to 3 minutes; cost
down 50 to 70 percent; projects run 9 to 12 months with ROI under one year.
• Benefit grid: low risk/non-invasive, accuracy, consistency, audit trail, productivity, reliability
(no sick days), right-shoring, scalability, retention, cross-industry, duration.
• Class grading of the 2016 brochure ten years on: the 93 percent claim is UNFALSIFIABLE
(not simply wrong); the 65 percent landed; retention missed; non-invasive partial. Bersin was
the primary grading witness; Novartis carried one claim and Gartner another. Vendor papers
get favorable framing — read critically, get references.
UiPath: HR Automation
• Internal story: 72+ automations in two years, about 30,000 hours saved per year (about 200
hours — five weeks — per HR employee). AI HR market: $3.89B (2022) to $17.61B (2027),
about 35 percent growth.
• Onboarding numbers: up to 20 percent of turnover happens in the first 45 days; strong
onboarding improves new-hire retention by 82 percent. Bertie (HSE's HR robot) does in one
hour what took the team five days.
• Other numbers: 75 percent of executives do not highly trust their data; 98 percent of
companies report a skills gap; culture connection cuts job-hunting by 55 percent. Best
practices: automate to innovate, make automation a core responsibility, and optimize before
you automate.
Recruiting Faster at SDL (Process-Analysis Case)
• Setup: SDL, Bangalore IT services, top three in India, $4B+ revenue, 80,000+ employees
(2009), recruiting 8,000 to 20,000 a year. Mohua Sen, new financial services recruitment
 head, reports to the VP of recruitment with a dotted line to business head Ganesh, who
measures her on exactly three things: quality, cost, and time. Every unfilled billable role
loses about $200 per day.
• Cycle: 80 to 90 days lateral. Funnel: 23 resumes → 13.8 test/interview → 5.52 technical →
1.82 HR discussion → 1 hire; that ratio is why $60 background checks stay late in the
process.
• Where time goes: join about 52 days (notice periods) and technical interview about 24.7 days
— both mostly WAIT time. Root bottlenecks: certified senior technical panels (weekend-
only, overloaded — Samir's complaint), late candidate documents, notice periods. Demand
peaks in Q2; 90 percent of requests come from seven of 14 units; duplicate PAS requests
(contract plus lateral) inflate the bench. PAS = People Allocation System; the PMO checks
internal talent first.

Unit 6: The Future of HR, Beamery, and IT Project Management

EY CHRO 2030 Market Insights
• Basis: 160+ executives, 15 sectors, 26 countries. The three headline stats: 85 percent say
strategic HR will be critical; 89 percent say HR must change notably; only 32 percent
currently have a talent advantage. The 89-vs-32 gap is the “critical opportunity”: urgency is
near universal, readiness is not.
• Three competencies: business, technology, humanistic. Four transformation areas in order:
the definition of work, the workforce, the business, the HR function. Four technologies:
AI/ML (prediction and decision support), generative AI (content creation), RPA (rules-based
execution, no judgment), advanced analytics (insight and measurement) — the classic
confusion is gen AI vs. RPA: one creates new content, the other repeats a defined process.
• Workforce numbers: global talent shortage projected to exceed 85 million people by 2030;
70 percent report urgent need for new skills; mobility — 90 percent see the benefit of
alignment, 70 percent of those cannot achieve it. Two workforce focus areas: shift from roles
to skills; improved employee experience.
Gartner: Build an HR Operating Model That Succeeds in the AI Era (Nieberg)
• The 30/70 rule: AI transformation depends 30 percent on the technology and 70 percent on
redesigning processes, workflows, and roles. The diagnosis pairing: 95 percent of CHROs
had adopted AI by end of 2025, yet nearly three quarters report only moderate or minimal
returns — tools without redesign.
• By 2030, 50 percent of current HR activities will be AI-automated or performed by AI
agents; AI augments 100 percent of HR tasks and performs up to 50 percent. 82 percent of
HR leaders plan agentic AI within 12 months. Opening line: HR must reinvent itself or be
dissolved; biggest risk of inaction is irrelevance, not lag.
• Three pillars (before → after): (1) transactional HR operations → digital HR solutions and
delivery, AI agents as the front door, ticket-based to ANTICIPATORY self-service; (2) fixed
HR business partners → dynamic PODS of strategic talent leaders (STLs), ratio moving from
 1:423 toward 1:800–1,200; (3) siloed centers of excellence → HR product designers building
hyperpersonalized solutions along horizontal value streams.
The Future of HR: What Needs to Be Reimagined by 2030 (LinkedIn Post)
• Know what it is: a practitioner OPINION post, opens with a Black Swan (Taleb) caveat —
cite as perspective, never evidence. Framing line: the next five years reward clarity, courage,
and creativity, not caution.
• Five focus areas: (1) Culture as a Product (CaaP) — users, pain points, features, feedback
loops; (2) workforce structure — plug-and-play squads of employees, freelancers, AI agents,
partners; missions over roles; (3) ethical AI guardianship — AI ethics boards, transparent
audit logs, AI literacy in leadership; (4) identity-driven talent systems — matching by
interests, strengths, potential instead of titles; (5) regenerative work design — energy
creation over burnout prevention; micro sabbaticals, asynchronous weeks.
Beamery: Using Skills and AI to Modernize HR (HBS Case)
• Founded 2013 in London by brothers Abakar and Sultan Saidov with Michael Paterson;
$223M raised; $1B unicorn valuation December 2022. Mission framing: talent is equally
distributed, opportunity is not (the passport lottery and the postcode lottery).
• Market pain stats: 77 percent of executives call hiring critical; 82 percent of CEOs do not
think they recruit the right people; only 7 percent believe they can retain talent; 40 percent of
HR leaders cannot build skills fast enough. Great Resignation: 88 million left jobs April 2021
to January 2023; top reason was lack of career development (ahead of pay).
• Skills-based hiring: 38 percent of workers hold four-year degrees and only 21 percent use
their major; Apple, IBM, and Google dropped degree requirements. “Jobs do not require
degrees, employers do.” Talent shortage forecast: $8.5 trillion in unrealized annual revenue
by 2030.
• Tech: knowledge graph with over 17 billion data points; TalentGPT (launched March 2023)
built on 20+ billion data points and a billion people's career transitions; explainable AI; no
automated employment decisions. NYC 2023 rule: yearly public bias audit, candidate notice,
alternative accommodations.
• Partnerships and proof: the ONLY endorsed Workday partner (two-way data access); also
SAP. Facebook was the first $100,000 customer (2016, at 14 employees); Amazon 2017.
Flux acquisition 2022 made it the first unified internal-and-external talent platform; Uber's
retention among underrepresented employees rose more than 200 percent. Internal-blindness
evidence: the construction client where 80+ percent of agency hires were rehired alumni; the
Goldman intern lost to headcount cuts and never rediscovered. Main competitor: Eightfold
(plus Avature, Phenom, Gloat). The case decision: focused specialist (higher satisfaction,
3.7–3.8 vs. 3.5) or multi-service player against Workday, SAP, Oracle.
IT Project Management (Video 6)
• Project = defined start and end; program = larger set of interlinked projects with evolving
objectives. Project portfolio management exists because demand for IT projects always
outstrips resource supply; five PPM capabilities: pipeline, resource, change, risk, financial
resource management. The PMO = the air traffic controllers of projects, created to fix the
inability to deliver on time, on budget, in scope.
• Technical debt = the price of short-term fixes plus the price of fixing them later. TCO covers
implement, operate, support/maintain, enhance/extend, decommission; commonly
overlooked: recurring license fees, support, maintenance, BYOD telecom usage. Cloud shifts
spending from CapEx to OpEx.
• Justification: never install technology for its own sake — always to solve a business problem.
Ward and Daniel six stages (stage four = identify the enabling ORGANIZATIONAL
changes). Three business change types: do new things, do things better, stop doing things.
Explicitness ladder high to low: financial, quantifiable, measurable, observable. HRIS benefit
types: cost reducing, cost avoidance, value enhancing (the sweet spot). Rule: buy to
standardize, build to compete. CapEx tools: benefit-cost analysis, NPV, IRR.
Competencies, Careers, and the 2018 Hype Cycle (Video 7 / Johansen Session)
• WEF number-one skill in both 2015 and 2020: complex problem solving; biggest riser into
2020: critical thinking (fourth to second). McKinsey 2030 categories: higher cognitive, social
and emotional, technological. Krogh's five competencies: enabling technologies, critical
thinking, digital ethics, skills for now and the future, communication.
• MIT Sloan's three new AI-era jobs: explainers, sustainers, trainers. Driver's four career
concepts: linear (upward), expert (little movement, for life), spiral (lateral every 7 to 10
years), transitory (lateral every 3 to 5). Automation rule: anything codifiable into rules with
available data is a candidate for machines.
• Gartner 2018 clusters: Democratized AI, Digitalized Ecosystems (LEGO example), Do-It-
Yourself Biohacking, Transparently Immersive Experiences, Ubiquitous Infrastructure.
Current-Events Layer (Lecture 6)
• Mobley v. Workday expanded into a nationwide ADEA collective action covering applicants
40+ screened by Workday's tools — the vendor, not just the employer, is being held liable.
The mechanism alleged is proxies (career gaps and similar features standing in for age or
disability), not explicit bias.
• January 2026: class action against Eightfold AI alleging it operated as an unregistered
consumer reporting agency under the Fair Credit Reporting Act — a PRIVACY attack,
distinct from bias. Regulatory map: California ADS regulations (Oct 2025, under FEHA),
Colorado AI Act (annual impact assessments, transparency notices), Illinois 2026
amendments. EEOC position: “the algorithm did it” is not a defense.
• VR training is past pilots: market $10.96B in 2026; Walmart trained a million+ employees
(Pickup Tower training from 8 hours to 15 minutes, retention of content up 70 percent); Bank
of America 50,000+ through VR; Tyson 20 percent injury reduction; Indiana DCS turnover
53 to 19 percent. Agentic AI adoption skews large: 48 percent of large businesses vs. 25
percent midsize and 4 percent small.

Master Matching Bank

Person → Idea or Role
John McCarthy Coined AI, 1955, Dartmouth Conference
Bob Johansen Shape-shifting, liquid organizations; more distributed authority
Josh Bersin Great Reinvention of HR; assistants → agents → Superagents;
Chipotle example
Marc Benioff “Last CEOs who manage only humans” (Davos)
Jim Roth Salesforce Customer Success president; agents under contact
center managers
Brian Millham Salesforce COO; simultaneous AI hiring and layoffs
Michael Hammer BPR, 1990, “Don't Automate, Obliterate” (with Champy:
radical redesign)
Dwane Lay Eighth waste: Underused people; approvals are always waste
Ed Beverly Gemba, as-is mapping; process mapping, seven wastes, root
cause analysis
W. Edwards Deming People don't fail, systems fail; “In God we trust, all others must
bring data”
Cathy O'Neil Weapons of Math Destruction; algorithms are opinions
embedded in code
Shoshana Zuboff Surveillance capitalism; people as raw material
Eli Pariser Filter bubbles
Derek Mobley Plaintiff suing Workday over AI screening bias (Feb 2023)
Dave Duffield and Aneel Ex-PeopleSoft executives who founded Workday (2005)
Bhusri
Saidov brothers Founded Beamery (2013, London) with Michael Paterson
Peter Cappelli and Anna HR Goes Agile
Tavis
Tomas Chamorro- Set Your Team Up to Collaborate with AI (five considerations)
Premuzic
Ben Dattner et al. Legal and Ethical Implications of AI in Hiring
Davenport, Harris, Competing on Talent Analytics; six uses; DELTA
Shapiro
Lynn Roger (BMO) “Speed is the new business currency”
Diane Gherson (IBM) Peer feedback in a network context
Jennifer Hyman (Rent the Rolled bonuses into base pay; “untangling the two”
Runway)
Joe Cava Google VP of Data Centers; “herder of cats”
Mohua Sen and Ganesh SDL recruitment head and the business head measuring
quality, cost, time
Chaubey and Sharma Apturja HR head choosing data over instinct; the MBA analyst
Trees Segers Novartis global HRIS program manager
Dan Vasella Novartis CEO, medical doctor
Hanne Nieberg Gartner AI-era HR operating model (30/70 rule)
Nassim Taleb The Black Swan (cited in the 2030 LinkedIn post)
Demis Hassabis DeepMind co-creator; AlphaGo's transferable algorithms

Number → Meaning
30/70 Gartner: AI success is 30 percent technology, 70 percent
operating-model redesign
85 / 89 / 32 EY: strategic HR critical / HR must change / actually have a
talent advantage
95 percent and three CHRO AI adoption by end of 2025; share reporting weak
quarters returns
50 percent by 2030 Share of HR activities AI-automated or agent-performed
(Gartner)
1:423 → 1:800–1,200 HRBP ratio to strategic talent leader ratio (Gartner pods)
30 to 40 percent / 60 Bersin: automatable HR jobs / five-year growth in HR postings
percent
100:1 → 300–400:1 Employee-to-HR staff ratio shift (Bersin)
93 / 65 / 65 / 20–35 EY RPA: repetitive time / automatable processes / cost saving
vs. offshore FTE / overall saving potential
2 hours → 3 minutes EY RPA processing time drop
72 / 30,000 / 200 UiPath: automations / hours saved yearly / hours per HR
employee
20 percent in 45 days / 82 Early turnover share / retention lift from strong onboarding
percent (UiPath)
80–90 days / $200 per day SDL lateral cycle / cost of an unfilled billable role
23 to 1 SDL resumes sourced per hire
52 / 24.7 SDL days for join / technical interview (biggest wait blocks)
79 percent Executives rating agile performance management a high
priority (Deloitte 2017)
20 → 46 percent / 3,000 J&J pilot manager participation / pieces of feedback
54 → 96 percent / $4M Jerry case inquiry response improvement / projected annual
savings
40 percent Recruiter time savings from AI (Chamorro-Premuzic)
70 percent of Fortune 500 Companies using Workday tools
77 / 82 / 7 / 40 Beamery pain stats: hiring critical / wrong hires / believe they
can retain / cannot build skills fast enough
88 million People who left jobs in the Great Resignation window
(Beamery case)
$8.5 trillion Unrealized annual revenue from the talent shortage by 2030
85 million Projected global talent shortage in people by 2030 (EY)
17 billion / 20 billion Beamery knowledge graph data points / TalentGPT training
data points
14.4 percent / $26.3B / 62.1 Cloud HCM market growth 2023 / size / Leaders' revenue
percent share
0.1 percent = $100K Best Buy: engagement gain to store annual income
10–15 percent / 20 percent Orgs fully exploiting big data / their outperformance (Gartner,
Intel video)
5 exabytes / 2.7 zettabytes All data through ~2003 / global data in 2012
455 → 545 Apturja resignations FY18 to FY19
CHF 78 million Novartis global HRIS approval (September 2002)
60/30/10 → 20/60/20 Novartis HR time: admin/consulting/strategy, before and after
full transformation
2 and 70 percent vs. 1 and Novartis ready-now successors and internal fill: target vs. 2003
42 percent reality
18 to 24 months Moore's Law doubling period
104 / 51 / 5 SAP core HR countries / SAP payroll localizations / Workday
payroll countries
Framework → Components
Designing for Dialogue Task analysis, interaction protocols, feedback loops (Baier)
WINS Words, Images, Numbers, Sounds (Baier)
DELTA Data, Enterprise, Leadership, Targets, Analysts (Davenport)
TIM WOOD U Transport, Inventory, Motion, Waiting, Over-processing, Over-
production, Defects, Underused people
DMAIC Define, Measure, Analyze, Improve, Control (Six Sigma)
Scott Morton model Strategies, business processes, structure, individual skills and
roles, technology (1991)
Gartner analytics ladder Descriptive, diagnostic, predictive, prescriptive
Hype Cycle stages Innovation Trigger, Peak of Inflated Expectations, Trough of
Disillusionment, Slope of Enlightenment, Plateau of
Productivity
Agile's six HR change Appraisals, coaching, teams, compensation, recruiting, L&D
areas (Cappelli and Tavis)
Chamorro-Premuzic's five Augmentation strategy; output over input; AI-resistant skills;
mid-level managers; experimentation
EY four transformation Definition of work, workforce, business, HR function
areas
EY four technologies AI/ML, generative AI, RPA, advanced analytics
Gartner three pillars Digital HR delivery; pods of strategic talent leaders; COEs as
HR product designers
2030 LinkedIn five areas Culture as a Product, plug-and-play squads, ethical AI
guardianship, identity-driven talent, regenerative work design
SHRM five keys Change HR with tech; employee self-reliance; communicate
monitoring; more data more answers; appearance and
functionality
SHRM five trends Social networking; compliance expansion; renting over buying;
BI and dashboards; transparency vs. privacy
J&J five conversations Goal setting, career discussion, midyear review, year-end
appraisal, compensation review
Waterfall seven steps Planning, analysis, design, development, testing,
implementation, maintenance
Ward and Daniel stage Identify enabling organizational changes (benefits come from
 four change, not software)
Explicitness ladder Financial, quantifiable, measurable, observable (high to low)
MIT Sloan three jobs Explainers, sustainers, trainers
Driver's four career Linear, expert, spiral (7–10 yrs), transitory (3–5 yrs)
concepts
Copeland's intelligence Generalization learning, reasoning, problem solving,
factors perception, language understanding
Krogh's five competencies Enabling technologies, critical thinking, digital ethics, skills for
now and future, communication

Vendor / Company → Placement or Claim to Fame

MQ Leaders Dayforce, Oracle, SAP, UKG, Workday
MQ Challengers ADP, Darwinbox
MQ Niche Players Cegid, Cornerstone, Infor, Yonyou (Visionaries: none)
Workday 70 percent of Fortune 500; best customer experience in MQ;
defendant in Mobley bias suit; Beamery's only endorsed
partner; NASCAR's HCM
Eightfold Beamery's main competitor; second Bersin lawsuit; Jan 2026
FCRA privacy class action
IBM Deep Blue vs. Weak AI (chess) vs. middle-ground AI (percent-confidence
Watson reasoning)
AlphaGo (DeepMind) First proven non-expert system
Toyota vs. Motorola Lean and Kaizen vs. Six Sigma
Chipotle Bersin's proof that AI hiring and scheduling speed raises
revenue
Wiley / OpenTable Agentforce proof points: $230K saved and 40 percent
improvement / shorter, better calls
Ascendify / HackerRank Agile-recruiting software vendors (GE, IBM, Cisco) / IT
recruiting screening tool
Telehouse North / 60 London main hub / largest east coast hub / LA hub
Hudson / One Wilshire

True/False Trap List (Built from Real Quiz Wording)

• Agile is a deductive process. FALSE — iterative.
• ERPs cannot be used to foster social responsibility. FALSE — they can.
• ERPs like SAP or PeopleSoft are best-of-breed applications. FALSE — they are integrated
suites, the opposite of best of breed.
• Systems of engagement are where enterprise data is stored, used primarily by internal
employees. FALSE — that describes a system of record.
• A relational database is primarily used to manage unstructured data. FALSE — structured;
unstructured goes to NoSQL.
• [Link] is an example of an IP address. FALSE — it is a URL; DNS translates
it to an IP.
• The internet is the world's largest client-server network. TRUE.
• A webpage request can take multiple paths and be reassembled in about a second. TRUE.
• A network packet is an unformatted unit of data. FALSE — formatted.
• SaaS refers to software hosted in the cloud; [Link] is an example. BOTH TRUE.
• All hackers are unethical. FALSE — white and gray hats exist.
• Any smart wireless-enabled appliance is a security threat. TRUE.
• Ransomware deletes or renames files. FALSE — it encrypts them.
• The Gartner Magic Quadrant presents strengths and weaknesses of individual vendors.
TRUE — previously missed; strengths and cautions per vendor.
• Upward feedback is valued in agile organizations. TRUE.
• Implementing a new HR system offers the opportunity to reexamine how a process is done.
TRUE (SHRM).
• New technology can be used to change HR work processes. TRUE (SHRM).
• AI can target training by matching backgrounds, career trajectory, and completed programs.
TRUE (IBM, in HR Goes Agile).
• Digital ecosystems: interdependent actors sharing standardized platforms for mutual benefit.
TRUE — definition verbatim.
• Johansen's shape-shifting organization means less distributed, more centralized authority.
FALSE — more distributed, less centralized.
• Target's pregnancy insight was descriptive analytics. FALSE — predictive.
• RPA replaces existing HR and payroll software. FALSE — non-invasive; it sits on top.
• RPA is a form of real AI. FALSE per lecture — rules-based execution, no judgment.
• Deep Learning contains Machine Learning, which contains AI. FALSE — the nesting runs
the other way.
• Most slow processes are dominated by work time. FALSE — delay/wait time dominates
(cycle time = process + delay).
• Superagents are AI tools that boost individual productivity. FALSE — they own an entire
end-to-end process.
• E-recruiting increases applicant quality. FALSE — volume, not quality.
• The Novartis HRIS was primarily an IT project. FALSE — the enabler of HR's
transformation to strategic partner.
• Gen AI should be treated as a traditional form of automation. FALSE — an assistive agent
that gets smarter over time (Baier).
• It is lawful to act on protected traits if they were inferred by an algorithm rather than asked
directly. FALSE — unlawful no matter how learned or inferred.
• In the Mobley case, only the employer faces liability. FALSE — the pivotal argument is
VENDOR liability.
• Beamery makes automated employment decisions. FALSE — explainable AI, no automated
decisions.
• The 2030 LinkedIn piece is peer-reviewed research. FALSE — an opinion post with a Black
Swan caveat.
• Gartner says AI transformation success mostly depends on picking the right technology.
FALSE — 30 percent technology, 70 percent redesign.

Fast Recall (Final Pass)

• Unit 1: ERP one database; Hype Cycle five stages (trigger, peak, trough, slope, plateau);
record vs. engagement; SaaS/PaaS/IaaS; Moore's Law 18–24 months; packet =
formatted, multiple paths, ~1 second; Johansen = more distributed; Superagents own
end-to-end processes; Roth → contact center managers own agents; Millham →
displacement AND creation.
• Unit 2: GIGO; four Vs; relational = structured; warehouse vs. operational; ETL; lake =
consistency problem; descriptive-diagnostic-predictive-prescriptive; Target = predictive;
AI ⊃ ML ⊃ DL; Watson = middle ground; AlphaGo = first non-expert; DELTA; Project
Oxygen 8 behaviors; Apturja = performance up, retention down, data over instinct.
• Unit 3: integrated vs. best of breed (ERPs are NOT best of breed); top buyer requirement
= cloud-based; Scott Morton five boxes; don't pave the cow path; ransomware encrypts;
DDoS non-intrusive; MQ Leaders = Dayforce, Oracle, SAP, UKG, Workday, no
Visionaries; MQ shows vendor strengths and cautions = TRUE; SHRM trends vs.
challenges; Novartis CHF 78M, 60→20 admin, don't web-enable bad processes.
• Unit 4: agile = iterative; six change areas, “All of these”; post-WWII = lifers, rules-
based, long view (NOT minorities); upward feedback True; Chamorro-Premuzic five
considerations, never pick the dehumanizing option; WINS = Baier; Designing for
Dialogue three components; Jerry 54→96 percent, $4M; adverse action unlawful
however inferred; Mobley: three statutes, disparate impact, EEOC amicus, vendor
liability; O'Neil, Zuboff, Pariser.
• Unit 5: cycle time = process + delay; TIM WOOD U (Lay's Underused people); Gemba;
Hammer and Champy; Kaizen/Lean Toyota, Six Sigma Motorola, DMAIC; RPA non-
invasive; EY 93/65/65/20–35, 2hrs→3min; 93 unfalsifiable; UiPath 72/30,000/200; SDL
80–90 days, 23:1, $200/day, wait time dominates.
• Unit 6: EY 85/89/32, 85M shortage, roles→skills; Gartner 30/70, 95 percent adoption vs.
weak returns, three pillars, 1:423→1:800–1,200; LinkedIn five areas (CaaP, squads, AI
guardianship, identity, regenerative) = opinion; Beamery 2013 London, 77/82/7/40, only
Workday partner, TalentGPT, no automated decisions, NYC bias audit; PMO = air
traffic controllers; buy to standardize, build to compete; explicitness ladder; WEF #1 =
complex problem solving.