My talk on AI for Human Resource Management at the Faculty Development Programme conducted by Department of Management Studies MVGR College of Engineering

1. AI for HRM
Gopi Krishna Nuti
Data Science Manager, Autodesk (gopi.nuti@autodesk.com)
Vice President, MUST Research (vp@must.co.in)

2. What is AI?
A study of how to make computers do things which, at the
moment, people do better.
Donald Knuth, Winner of Turing Award (considered as Nobel of Computer Science)

3. Demystify the
terminology

4. Where is
decision
making?
How to commute to railway station
Deciding action on the day before
exam
Deciding on selling on stock
market
Machine Learning
using Statistics
•Calculate the likelihood of delay when
travelling by Car?
•Calculate the likelihood of delay when
travelling by city bus?
•Calculate the likelihood of running out
of pocket money?
•Calculate the probability of
failing if I go to movie?
•Calculate the probability of
passing if I study all night?
•Calculate the probability of
market gaining x points
tomorrow
•Calculate probability of market
gaining y points tomorrow
•Calculate probability of market
gaining z points next week
Machine Learning
using Linear algebra
a.k.a Deep Learning
•Calculate the likelihood of delay when
travelling by Car?
•Calculate the likelihood of delay when
travelling by city bus?
•Calculate the likelihood of running out
of pocket money?
•Calculate the probability of
failing if I go to movie?
•Calculate the probability of
passing if I study all night?
•Calculate the probability of
market gaining x points
tomorrow
•Calculate probability of market
gaining y points tomorrow
•Calculate probability of market
gaining z points next week
Artificial Intelligence •Should I take a taxi or city bus?
•Should I go to a movie or study
all night?
•Should I sell stock tomorrow or
next week?

5. AI in Human Resource
Management

6. What is done
by Humans in
HRM?
• Strategic HRM
• Human Resource Planning
• Planning
• Recruitment
• Placement
• Training and Development
• Performance Management and Appraisal
• Coaching and Talent Management
• Policy communications
• Compensation
• Basic Pay, Performance Pay, Benefits and Incentives
• Ethical aspects
• Employee Safety, Health and Incident management
Traditionally, HR was thought
to be an all-human field.

7. Current state
of AI
adoption
• 64% of People Trust a Robot More Than Their Manager
Acceptance
• 50% of workers currently using some form of AI in 2019
compared to 32% in 2018
• 65% of workers are optimistic, excited and grateful about
having robot co-workers
• Nearly 25% report having a loving and gratifying relationship
with AI at work.
• 32 percent of men have optimistic view of AI vs. 23 percent of
women.
Penetration
Source: https://www.oracle.com/corporate/pressrelease/robots-at-work-101519.html

8. Current state
of AI
adoption
• 64% of workers trust a robot more than their manager
• 50% have turned to a robot instead of their manager for advice.
• 82% think robots can do things better than their managers.
Acceptance
• Providing unbiased information (26%)
• Maintaining work schedules (34%)
• Problem solving (29%)
• Managing a budget (26%).
Where are robots doing better?
• Empathy (45%)
• Coaching (33%)
• Creating a work culture (29%).
Where are humans better?
Source: https://www.oracle.com/corporate/pressrelease/robots-at-work-101519.html

9. Why is AI adoption growing?
24/7 AVAILABILITY AUTOMATION PERSONALIZED
COMMUNICATIONS
REAL-TIME DATA.

10. AI for Talent Acquisition
Common Challenges
• Time consuming
• Subjective
• Reactive
How is AI helping - Find the right talent at just the right time
• Recommend right candidate for the job. Go beyond a simple search for key terms.
• Recommend right jobs to candidate
• Predict candidate performance
• Calculate the likelihood of accepting the offer, performance outcomes and expected tenure.
Screening and Interview
• Automated interactions with the help of a digital assistant
• Scheduling, re-scheduling, cancelling the interviews,
• Reminders, recommending preparation material, sending feedback
Selecting and Offering
• Compare candidates with each other and against benchmarks
• Create individualized offers
• Anticipate candidate behavior regarding the offer

11. AI for Talent Management
Common Challenges
•Employee retention
•Passive career development
•Traditional succession planning
•Rigid, undifferentiated learning
•Compensation expectations
How is AI helping
•personalize career development
•optimize succession planning
•close skills gaps
•Steer compensation strategy
Career development
•Personalized career development recommendations for Employees
Succession Planning
•Identify employees who are likely to leave
•Identify most capable successors
Compensation
•Provide market insights
•Increase recruiting efficacy

12. AI for Learning and Development
Common Challenges
•One size fits all approach of L&D systems
How AI is helping
•Personalized learning
•Optimize administration for L&D managers

13. https://www.gartner.com/smarterwithgartner/3-
trends-in-the-gartner-hype-cycle-for-human-capital-
management-2020

14. Case study 1 – Resume Parser
Resumes are structured in myriad and
complex ways.
Recruiters would prefer to have a
consistent way of reading the
information to compare and contrast
candidates.
Solution is resume parsers
Ex: SeekOut, Resume Parser by Affinda, DaXtra
Parser, HireAbility ALEX Resume and CV Parser,
Rchilli and many more

15. S. No Name Contact Qualification Years of
experience
Current CTC Location

16. Case Study 2 – Chat bots
Recruiters at FirstJob could not pay
attention to interviews and closing
offers.
SOLUTION:
A chatbot named Mya.
Mya can communicate with numerous
candidates at once and ask them pre-
screening questions.
Half of a recruiters’ job is automated
as this chatbot answers candidates’
questions and alerts them when a job
position is filled up.
RESULT:
An increase in recruiters’ efficiency by
38%.
Up to 75% of the qualifying process is
automated.
An increase in job engagement
process by 150%.
https://www.rchilli.com/blog/top-3-case-studies-showing-ai-power-in-simplifying-recruitment

17. Case study 3 – Predicting Employee Attrition
A much-respected data set from IBM is publicly available for building our own models for
this.
Variable Meaning Levels
Age Age of the employee
Attrition Whether the employee left in the previous year or not
BusinessTravel How frequently the employees travelled for business purposes in the last year
Department Department in company
DistanceFromHome Distance from home in kms
1 'Below College'
2 'College'
3 'Bachelor'
4 'Master'
5 'Doctor'
EducationField Field of education
EmployeeCount Employee count
EmployeeNumber Employee number/id
1 'Low'
2 'Medium'
3 'High'
4 'Very High'
Gender Gender of employee
1 'Low'
2 'Medium'
3 'High'
4 'Very High'
JobLevel Job level at company on a scale of 1 to 5
JobRole Name of job role in company
1 'Low'
2 'Medium'
3 'High'
4 'Very High'
MaritalStatus Marital status of the employee
MonthlyIncome Monthly income in rupees per month
NumCompaniesWorked Total number of companies the employee has worked for
Over18 Whether the employee is above 18 years of age or not
PercentSalaryHike Percent salary hike for last year
1 'Low'
2 'Good'
3 'Excellent'
4 'Outstanding'
1 'Low'
2 'Medium'
3 'High'
4 'Very High'
StandardHours Standard hours of work for the employee
StockOptionLevel Stock option level of the employee
TotalWorkingYears Total number of years the employee has worked so far
TrainingTimesLastYear Number of times training was conducted for this employee last year
1 'Bad'
2 'Good'
3 'Better'
4 'Best'
YearsAtCompany Total number of years spent at the company by the employee
YearsSinceLastPromotion Number of years since last promotion
YearsWithCurrManager Number of years under current manager
Job Involvement Level
JobInvolvement
JobSatisfaction
Education Education Level
EnvironmentSatisfaction Work Environment Satisfaction Level
RelationshipSatisfaction Relationship satisfaction level
Work life balance level
WorkLifeBalance
Job Satisfaction Level
Performance rating for last year
PerformanceRating

18. NLP for
Employee
Satisfaction
• Easy to create very powerful chatbots
• Conversational trees can be maintained to provide a
near-human experience
Chatbot for answering queries
• AWS (Lex, Polly), Google Dialog Flow, MS Azure’s Bot
service and Bot Framework.
• Rasa for Open-source lovers
Software
Source: https://www.oracle.com/corporate/pressrelease/robots-at-work-101519.html

19. NLP for
Employee
Satisfaction
• Sentiment analysis for analysing survey responses
• Traditionally, surveys are multiple choice questions
On multiple occasions, when I approached my supervisor, he was helpful.
But on a few occasions, I felt he can be better approachable. His cell
phone habits are terrible so, we can’t rely on that. But whenever he is in
office, he is very approachable and always patient and helpful.

20. NLP for
Employee
Safety
• Help desk tickets contain a wealth of information.
• Identify patterns and risks based on the text
Identifying health risks
Source: https://www.oracle.com/corporate/pressrelease/robots-at-work-101519.html

21. Food for
thought
Source: https://www.oracle.com/corporate/pressrelease/robots-at-work-101519.html
• Can we predict which employee’s US Visa
application is likely to be approved or
rejected?
• US L1 A/B applications have very strict and
confidential criteria.
• A lot of information is asked by the consulate
• Can we tabulate all that information and
identify the successful criteria?