Generative AI for Enterprises – Don’t Hype, Adopt Simply
Generative AI for Enterprises – Don’t
Hype, Adopt Simply
By: Subham Sarkar
There will be very few who will disagree that Artificial Intelligence (AI) is the biggest disruption after Cloud in the world of IT and business models at large globally.
AI has been evolving over many
years now with an ever-increasing set of terminologies:
a)
Monikers like Intelligent AI, GenAI, Agentic AI, etc.
b)
Model Types like Machine Learning, Supervised Learning,
Unsupervised Learning, Deep Learning
c)
Branches
like Computer Vision, Fuzzy
Logic, Expert Systems, Robotics, Machine Learning, Neural Networks/Deep
Learning, Natural Language Processing
d)
Models
& Assistants (Chat BOTs) like ChatGPT, Gemini, Deep Seek, Claude, Manus, etc.
To add to these AI-related
vocabulary, there are also 7 commonly agreed Types of Artificial
Intelligence (Source Credit: Sunny Betz)
- Narrow AI: AI designed to complete very specific actions; unable to
independently learn.
- Artificial General Intelligence: AI designed to learn, think and perform at similar levels to
humans.
- Artificial Superintelligence: AI able to surpass the knowledge and capabilities of humans.
- Reactive Machine AI: AI capable of responding to external stimuli in real time;
unable to build memory or store information for future.
- Limited Memory AI: AI that can store knowledge and use it to learn and train
for future tasks.
- Theory of Mind AI: AI that can sense and respond to human emotions, plus
perform the tasks of limited memory machines.
- Self-Aware AI: AI that can recognize others’ emotions, plus has sense of
self and human-level intelligence; the final stage of AI.
While AI has been evolving rapidly with huge progress wrt technology and
theory, the practical aspects of AI (How to Use AI) however has mostly
not landed firmly on the ground yet. If one asks a CXO or a CIO about AI,
almost all will enthusiastically say “YES, we are Using AI”, but when probed
further most will get into generalizations without getting into the specifics
of What, Why, How, etc.
It’s like a “déjà vu” moment with what happened with “Digital
Transformation” – everybody started talking about it, but there was no
common articulation or set process of “How to Do Digital Transformation”.
The fact remains that no matter how disruptive or revolutionary a new
technology may be, until and unless it adds demonstrable, repeatable and
significant value to businesses (through business cases, ROI, etc.) wrt
Business Growth/Transformation/Modernization, it will just end up being another
hype or jargon.
In this article, I (while upfront declaring that I am no expert on
AI) will attempt to filter out all the hype, hoopla and noise about AI/Gen
AI - while laying out my viewpoint about a simple, common-sensical and
business-contextual “2-Steps” way of adopting this useful technology evolution
by enterprises, irrespective of their size, geography spread and industry
vertical. I will also elaborate on a potential low-hanging fruit which can be
the first use case for Gen AI adoption.
STEP 1: PROOF OF THE PUDDING (The Appetizer)
Identifying the first use case
(aka low hanging fruit) for GenAI adoption while ticking off the
following considerations:
o
It will address a business
problem which is visible and organisation-wide
o
It will deliver significant
business benefits
o
It will entail least/a low risk
of regular business disruption
o
It will ideally be in the
control/domain of IT department
o
It will not require existing
technology/platforms/tools to be replaced or disturbed
o
It will align with the existing
& future overall technology direction, architecture, landscape and Cybersecurity
posture
o
It will have the attention of
the Top Leadership/Management
Once a use case is identified
applying the aforesaid criteria, a thorough Business Case needs to be documented
and approved through the organization’s approval/stage gate processes. The
business case needs to cover the following:
a)
Business problem to be solved
b)
Quantification of the business
problem
c)
Cost of adoption/implementation
d)
Quantification of business
benefits & also Qualitative benefits expected post adoption
e)
ROI expected
STEP 2: ROLLOUT ROADMAP & EXECUTION (The Main
Course with Dessert)
With the first proof point
successfully delivered (after business case re-validation, post the
implementation), next logical step would be to:
1)
Identify the next set/list of
use cases and create a rollout roadmap based on priority/business case/stage
gate approvals
2)
IT Team can work with other
business functions in the organization like Finance, Procurement, Supply Chain,
Manufacturing, Sales & Marketing, Distribution, HR & Administration,
etc. for identifying and detailing these use cases
3)
Create Business Cases for each
Use Case listed
4)
Set up a “AI Centre of
Excellence” (CoE) to rollout these GenAI use cases across the organization
5)
IT Team can also take external
help from consultants or tech service providers for this exercise, who can
bring in expertise & partnership models like CoE Consulting & Staffing,
IP Co-Creation, GTM for CoE as a Profit Centre, etc.
The goal and outcome would be to
effectively enable the enterprise’s Business Transformation initiatives with Digital
Transformation through new tech adoption
IDENTIFYING THE “LOW HANGING FRUIT” (First GenAI Use Case)
WHAT
o
IT Support Tickets
WHY
o
IT Support is a key function of
IT as an enabler for business transformation
o
It impacts both the revenue and
cost sides of any enterprise irrespective of size, geography spread and
industry vertical
o
From an IT Service Delivery
perspective covering Infra, Apps, Integration, etc. it impacts the daily work
lives and productivity of internal stakeholders including end users across various
business functions, IT support teams, outsourced vendor support teams) and also
external stakeholders in cases where IT platform(s) access is given to
customers and vendors
o
IT Support, while "Keeping
the Lights On", also gets a significant allocation of the IT Budgets,
ranging from 30%-40% in most manufacturing & distribution-oriented
enterprises
o
In most enterprises, COTS applications
like ServiceNow, Fresh Service, Pager Duty, etc. and also homegrown legacy apps
are used as tools for IT Support, with a combination of multiple tools being
used is also common in enterprises
o
IT Support Tickets are primarily
of 2 types - Incidents and Enhancements, with Incidents comprising 60% - 80% of
the total tickets raised
o
In practical life, the quality
of enhancements work unfortunately also leads to new Incident support tickets
being added to the list
o
Incident Support Tickets are
commonly categorized wrt workflows/responsibilities like L1, L2, L3, L4 and
also wrt complexity/severity like P1, P2, P3 etc.
o
Metrics like Ticket Trends,
MTTA, MTTR, SLA Violations, etc., are commonly used to govern the IT Support
service delivery
o
In most enterprises, an
elaborate support team structure including internal and outsourced vendor
support is in place
o
In many cases, reinventing the
wheel, i.e., solving "known incidents" are a major workload for these
support teams, which ideally should be an end user self-service than a new
ticket raised
HOW
o
In the first phase of
adoption ("BOT Agent with Human in Loop"), the BOT
solution will enable/guide end user self-help for repeats of all
"known" and "similar" support tickets.
o
In the next phase of adoption
("BOT Agent Only/Agentic
AI"), the BOT will execute "self-heal" or "auto-care" for repeats of all "known" and
"similar" support tickets.
o
Further, as & when “new/unknown”
tickets are solved by the Support teams, the BOT will just need to get
trained on these new incident resolutions
BENEFITS
o
Lesser No. of Support Tickets
being raised - enhanced end user satisfaction and productivity
o
Lesser No. of Support Team
(internal & outsourced vendor) required, with them being redeployed for
enhancements or other productive IT tasks/projects - reduced IT support
costs
o
While IT Support platforms like ServiceNow,
Fresh Service, Pager Duty, Azure DevOps, etc., evolve and may offer some
Analytics & AI Automation wrt Self Service, that would be a hybrid approach
and will entail significant integration costs
and additional licensing costs for giving access to all end users
o
Sitting on top of these
aforesaid multiple ticketing tools being used, the Gen AI solution will be like
a single common abstraction layer - a single touchpoint for end users
or support teams without the need to navigate across multiple tools
o
This solution can also be tailored
to align with the existing & future overall technology direction,
architecture, landscape and Cyber Security posture
o
It will entail least/a low risk
of regular business disruption
o
It will be in the control/domain
of IT department
o
It will not require existing
technology/platforms/tools to be replaced or disturbed
o
Unlike humans, the BOT can work
24x7x365 without fatigue and human errors
A Blog Series on New &
Emerging Technologies
Dated: March 22, 2025
Author: Subham Sarkar (https://www.linkedin.com/in/subham-sarkar-519b7114/)
- Strategic Advisor (Celebal Technologies, Agilitz Technologies,
Netlabs Global IT Services, Jumpstart Ninja Technologies)
Disclaimer:
The contents of this article are purely written in an individual capacity
based on the personal opinions of the author. The author does not claim to be a
SME on the topic covered in this article.
Data sources and image credits have been duly cited, as and where
applicable.
Other New & Emerging Technology Articles by
this Author:
https://industry4o.com/2024/02/13/gen-ai/
" GenAI - A Case of Running with the Hare and
Hunting with the Hounds ? "
https://industry4o.com/2022/08/23/generative-adversarial-networks-gans-an-ai-ml-based-future/
" Generative Adversarial Networks (GANs) - An
AI & ML based Future "
https://industry4o.com/2023/06/22/quantum-computing/
" Quantum Computing – This Hype is for Real!
”
https://industry4o.com/2023/04/12/india-worlds-digital-economy-leader/
"India - World's Digital Economy Leader"
https://industry4o.com/2023/01/14/future-of-transportation/
" Future of Transportation"
https://industry4o.com/2021/09/30/how-artificial-intelligence-is-gaining-greater-levels-of-trust/
" How Artificial Intelligence is gaining
greater levels of trust in elderly care"
https://industry4o.com/2021/06/18/all-you-need-to-know-about-sentiment-analysis/
"All you need to know about Sentiment
Analysis"
https://industry4o.com/2020/09/09/intelligent-automation-in-its-early-teens/
" Intelligent Automation in it's Early
Teens"
.jpg)
Comments
Post a Comment