Agile Ahmedabad 12-13 Sep 2025 | Data, AI & Human Ingenuity: Building Intelligent Delivery Ecosystems by Bhavesh Mehta

This document outlines how the fusion of data, AI and human creativity is shaping intelligent delivery ecosystems, strategies to harness technology and human ingenuity for smarter, faster and more adaptive outcomes.

1. Data, AI, and
Human Ingenuity
Building Delivery Ecosystems
Bhavesh Mehta
Chief Delivery Officer @ Apexon
September 12, 2025

2. 2

3. 3

4. 4

5. 5
Fiction to reality timeline of AI Technologies
JetsonsRosie The Robot Maid 1962
Star Trek Th Universal Translator 1966
Doctor Who Robot master at Chess 1977
Terminator Self Aware Weapons 1984
Back to the Future Smart Glasses and Smart
Watches 1989
A Space Odyssey HAL 9000 2001
Minority Report Brain Machine Interface 2002
AvatarExtraordinary Visual Experiences 2009
HER Digital Companion 2013
1997 IBM Deep Blue beats Garry Kasparov
2002 IRobotRoomba navigates rooms and obstacles in a
random pattern
2013Google Glass handles phone calls, texts, photos,
videos and maps
2015 Apple Watch Optical hear rate sensor capable of
spotting a heart condition
2019 Google Assistant Interpreter Mode translates 27
languages like in pixel buds
2021Metaverselaunched with AR and VR experiences
2022 Chat GPT capable of creating realistic conversational
experiences
40 years
9 years

6. 6
40% of Employees Report Increased Productivity from AI Adoption
Source: Upwork
72% of Executives Believe AI Will Be a Key Business Advantage
Source: CFO
95% of Fortune 500 Companies are Currently Utilizing AI in Some Capacity
Source: CyberNews
AI is Not the Future; It's the Business of Today
Why Artificial Intelligence? What’s the Big Deal?
More than Ever, Enterprises are Driving Speed,
Scale, and Smarter Decisions with AI

7. 7
VideoDemo
AI Capabilities in 2025

8. 8
So… What is AI Anyway?
“AI is when humans use data to train computers to think, learn, and solve
problems like a human would.”
Data
(The Fuel)
The core. None of this works without
vast amounts of high-quality data. It is
the fuel for the machine learning engine.
Machine Learning
(The Engine)
This is the most common type of AI today.
It’s the engine that learns from data to find
patterns and make predictions without
being explicitly programmed for every
scenario.
Artificial Intelligence
(The Big Idea)
It’s the broad concept of creating systems
that can perform tasks that normally
require human intelligence.
The Foundational Layers…
Examples
Voice Assistants Face ID Smart Home Devices Social Media Feeds Personalized Advertising Ride-Share Services

9. 9
Car Analogy: From BI to AI Maturity
Example:
“Speed 60 k/hr, gas
tank half full”
Example:
“Fuel efficiency dropped
because of city driving.”
Example:
“Based on traffic ahead,
arrival will be delayed 15
minutes.”
Example:
“Autonomous vehicles
sensing and driving without
human input”
Tells you what’s happening
right now
Shows what happened and
why
Predicts what will likely
happen
Predicts and acts
automatically
Business Intelligence
(BI Reports)
Analytics (Diagnostic
Reports)
Machine Learning
(Prediction)
Artificial Intelligence
(Action)

10. 10
Let’s Demystify the Buzz Words
ASPECT
Primary
Question
Focus
Key Output
Vehicle
Analogy
BUSINESS
INTELLIGENCE
(BI)
What happened?
Past & Present
(Descriptive)
Dashboards,
Reports
Speed, Fuel
Level, RPMs
ANALYTICS
Why did it
happen? What will
happen next?
Future (Predictive)
Forecasts,
Insights
GPS Navigation
MACHINE
LEARNING
(ML)
How can we learn
from data to
improve a task?
Pattern
Recognition
A trained
prediction model
Self-Improving
Engine
ARTIFICIAL
INTELLIGENCE
(AI)
How can we
automate complex
decisions?
Autonomous
Action
An intelligent
system
The Self-Driving
Car
GENERATIVE
AI
(GENAI)
How can we
create new and
original content
from data?
Content Creation,
Simulation,
Personalization
Text, Images,
Code, Video,
Designs, Music
The Creative Car
AGENTIC AI
How can AI act
independently to
achieve goals?
Goal-Oriented
Autonomy
Autonomous
Agents
Fully-Autonomous
Car

11. 11
Despite massive potential of AI in software development,
many companies are struggling to capture value at scale
Potential from GenAI for IT & Software Typical productivity gains
$900 B Productivity uplift across
product development and corporate IT
31% of global enterprise
software spend
Content Courtesy: McKinsey & Company

12. 12
1.Tech
ecosystem
and tooling
2.Operating
model
3.Talent
… and attention is on tools
but we need to fundamentally reshape skills and ways of
working
There is an abundance of AI SW
development tools…
but they only address a small piece of the problem
Full opportunity
<30% of
engineering time
spent writing code
Ideation Development Deployment
O
p
e
r
a
t
io
n
E
n
g
i
n
e
e
r
in
g
Produc
t
Design
Why are we not (yet) realizing full impact from AI in
software development?
Content Courtesy: McKinsey & Company

13. 13
Requirement
generation
Convert pain points and
ideas into detailed epics,
features, and stories
Design generation
Turn plain text into low-to
high fidelity mockups
& designs
Pain point synthesis
Identify, structure and
prioritize pain points from
customer interactions
(usage, feedback)
Code discovery
Scan and analyze code
base to provide better
context awareness and
identify modernization
opportunities
Technical requirements
generation
Generate system
architecture, entity
relationship diagrams,
and cloud architecture
Quality discovery
Examine code quality,
producing heatmaps and
metrics to identify areas
needing refactoring or
cleanup
Requirement to code
Generate accurate,
context-aware code from
user stories
Design to code
Create front-end to
backendintegration logic
based on designs
Prototype build
Rapidly build working
prototypes with front-&
back-end functionality to
test new ideas
Code generation (e.g.,
GitHub Copilot,
Cursor)
Generate, explain and
review in-line code and
generate Unit Tests,
directly within the IDE
Code review (e.g.,
CodeRabbit)
Review pull requests for
code quality and security,
streamline code review
with summaries & actions
Code translation and
migration
Translate codebases to a
new language or migrate
to a new version
Unit test generation
Automatically add unit
tests throughout the code
base based on code,
comments, and
requirements
Functional testing
Suggest and generate UI
tests along with code
changes
Non-functional testing
Predict changes in
availability / performance
based on changes to
codebase
Monitoring & Support
Triage, run root cause
analysis, and resolve
production issues based
on historical issues and
real-time monitoring
Feature usage
insights
Summarize user behavior
and suggests potential
features &
improvements
Performance -Cost
Optimization
Track tooling adoption,
retention, and impact
Monitor & OperateTest & ReleaseBuildViabilityDiscovery
Full opportunity
Most popular use case today Rapidly emerging solutions
We are now seeing a much more fundamental shift as AI
transforms the end-to-end product development lifecycle
Example AI use cases across the product development lifecycle
Content Courtesy: McKinsey & Company

14. 14
Building blocks of an AI-powered software transformation
Content Courtesy: McKinsey & Company

15. 15
Common Use-Cases in the Industry
SALES
•Score sales
leads
•Personalized
emails
•Conduct
outreach
•Forecast
revenue
MARKETING
& DESIGN
•Segment
customers
•Analyze
Sentiment
•Optimize
campaigns
•Product
designs
MARKETING
& DESIGN
•Segment
customers
•Analyze
Sentiment
•Optimize
campaigns
•Product
designs
DECISION
MAKING
•Financial
modeling
scenarios
•Business
strategy
scenarios
•A/B testing
DECISION
MAKING
•Financial
modeling
scenarios
•Business
strategy
scenarios
•A/B testing
SMART
FACTORY
•Generate
optimal
factory
layouts
•Detects
product
defects with
computer
vision
SMART
FACTORY
•Generate
optimal
factory layouts
•Detects
product
defects with
computer
vision
SUPPLY
CHAIN
RESILIENCE
•Optimize
shipping
routes
•Score
supplier risk
•Inventory
optimization
SUPPLY
CHAIN
RESILIENCE
•Optimize
shipping
routes
•Score supplier
risk
•Inventory
optimization

16. 16
Autonomous Delivery & Warehousing
ImpactofAI-PoweredRobots
AIroboticscreateapowerfulcollaborationbetweenhumanexpertiseandautomatedsystems.
20% Reduction
Infulfillmentcosts
40% Increase
Ordersprocessedperhour
99.8%Accuracy
PickingaccuracywithcomputervisionsystemsAmazon's Robotic Revolution
Amazon'swarehouseroboticsprogramincludesover520,000AI-
poweredrobotsworkingalongsidehumans,transforming
fulfillmentcenteroperations.

17. 17
The Human-AI Partnership

18. 18
The Human-AI Partnership →
Why human intelligence, intuition, and ethical judgment remainindispensable
It is vital for ensuring responsible and effective AI deployment. It emphasizes that human oversight and intervention are necessaryatvarious stages of
AI system operation, especially in complex or sensitive domains.
EthicalDilemmas&Bias
Humaninterventionis essentialtoidentifyand
mitigatebiasesinAItrainingdata,ensuringfairness
andequityinAI'sdecisions.
ComplexDecision-Making
Human nuance,commonsense,andcontextual
awarenessareirreplaceableforsubjectivejudgments
andunderstandingunquantifiablefactors.
CreativeInnovation
Truecreativityandbreakthroughinnovationstem
fromhumanimaginationandtheabilitytoconnect
disparateconceptsinunforeseenways.
Accountability
HumansbeartheresponsibilityforAIactionsand
impacts. Whenerrorsoccur,humanaccountabilityis
paramount forethicalandlegalcompliance.
EdgeCases&Ambiguity
Humansexcelathandlingnovelsituationsand
ambiguousdatathatAImodelsmayfailtoaddress
properly.
Human-AICollaboration
ThefutureofintelligentITdeliveryecosystemshinges
noton AIreplacinghumans,butonapowerful
partnershipwhereAIaugmentshumancapabilities.

19. 19
AI
Awareness
AI
Enablement
AI Adoption
Change management is key

20. 20
Call to Action: Academia & Students
Todrivethefutureofintelligentdeliveryecosystems,academiaandstudentsmusttakeactionnow.
The future of delivery ecosystems depends on your contributions.
Pursue Interdisciplinary
Research
Bridge computer science, business,
ethics, and social sciences
Address complex challenges in
intelligent delivery
Develop new algorithms and ethical
frameworks
Develop Future-Ready
Skills
Data science and artificial intelligence
expertise
Machine learning and systems thinking
Critical thinking and problem-solving
abilities
EngageinIndustry
Collaboration
Seek internships with industry partners
Complete capstone projects with real-
world applications
Establish academic-industry
partnerships

21. 21
Industryleadersmustchampionresponsibleinnovationinintelligentdeliveryecosystems.
"The future belongs to those who balance technology innovation with human values."
ChampionData-Driven
Decision-Making
Cultivate organizational culture that
values data insights
Use data for strategic planning
Make data accessibleto allemployees
InvestinWorkforce
Upskilling
TrainemployeesinAIcollaborationskills
Focus on critical thinking and problem-
solving
Elevate roles from transactional to
strategic
AdoptResponsibleAI
Principles
Implement ethical and transparentAI
practices
Ensure algorithmic fairness
Buildtrustwithstakeholders
Call to Action: Industry Professionals

22. 22
In Closing… Survival of the fittest
AI adoption is no longer optional—it’s a strategic necessity.
Like the smartphone, once a luxury but now indispensable, AI
is rapidly becoming the foundation for efficiency and
competitiveness.