SlideShare a Scribd company logo
A Fireside Chat moderated by
David Loshin
Affiliate Research Director, TDWI
President, Knowledge Integrity, Inc.
Senior Lecturer, University of Maryland
November 2, 2023
Supercharging AI with Data Enrichment
SPONSOR
2
President, Knowledge Integrity, Inc.
Senior Lecturer and Lead for External Relations,
University of Maryland
What we will talk
about today
• Setting the stage:
• Generative AI as a business imperative
• Data imperatives for AI model quality
• Discussion: The pivotal role of data
enrichment in training and fine-tuning
Generative AI models
Emergence of Generative AI
What is Generative AI?
• A subset of artificial intelligence that includes systems designed
to generate outputs such as images, music, text, or other forms of
media, based on its training data
• Learns from existing and generate new data that is consistent
with the original data set
• Generative AI systems that have been trained on billions of
parameters use prediction to create new instances of data in
response to provided prompts
• Large Language Models (LLMs) are a type of Generative AI that
have been trained on massive amounts of content
Ensuring Trustworthy & Appropriate Results
Data
volumes
Data
quality
Data
access
• Issues include:
– Bias
– Privacy
– Ethical concerns
– Legal concerns
– Hallucinations
Data Enrichment & LLM Training
• Improving the utility of data through appending and integration of
relevant content from additional sources
• Enrichment is used for
– Refining contextual nuances
– Improving fidelity of prompt responses
– Improve pattern recognition to reduce probability of hallucinations
– Improve interpretability of results
Supercharging AI with Data Enrichment
The leader in data integrity
Our software, data enrichment products and
strategic services deliver accuracy, consistency, and
context in your data, powering confident decisions.
of the Fortune 100
99
countries
100 2,500
employees
customers
12,000
Brands you trust, trust us
Data leaders partner with us
10
AI initiatives succeed with trusted data
of leading
businesses have
ongoing investments
in artificial
intelligence
91%
From Noise to Brilliance: Supercharge AI with Data Enrichment
Algorithms
Data
Modeling
Large
Language
Models
Deep
Learning
Hyperparameter
Tuning
Training
Data
Retrieval
Augmented
Generation
Supervised
Learning
Natural
Language
Processing
Bias
and
Fairness
Artificial
Intelligence
Feature
Engineering
Neural
Networks
Chatbots
Machine
Learning
Data
Mining
11
Source: NewVantage
For trusted data,
you need data integrity
Data integrity is data with maximum
accuracy, consistency, and context for
confident business decision-making
Data
Integrity
From Noise to Brilliance: Supercharge AI with Data Enrichment
12
What is data
enrichment, exactly?
13
It’s the process of enhancing your data by
appending relevant context from additional
sources – improving its overall value,
accuracy, and usability.
From Noise to Brilliance: Supercharge AI with Data Enrichment
Trusted third-party data at a global scale
Addresses &
Property
Verified and validated address and
property data for map display and
analytics
Boundaries
Administrative, community, and
industry-specific boundaries for data
enrichment and territory analysis
Demographics
Demographic and consumer context
data for better understanding people
and behavior
Points of
Interest
Detailed business, leisure, and
geographic features for location
and competitive intelligence
Streets
Robust street-level data for mapping,
analysis, routing, and geocoding
Risk
Natural hazard boundaries related to
flood, fire, earthquakes, and weather
14
Expertly curated datasets containing thousands of attributes for faster, confident decisions
From Noise to Brilliance: Supercharge AI with Data Enrichment
15 From Noise to Brilliance: Supercharge AI with Data Enrichment
Purchases &
Shopping
Building & Parcel
Boundaries
Lifestyles
PreciselyID
School Rankings
Points of Interest
Addresses Population
Property Attributes
Weather
Natural & Manmade
Hazards
Travel Time
Administrative
Boundaries
Land & Property Consumer Environment
Data enrichment can be easy with the right tools
A unique identifier for every address that doesn’t change, and other methods for appending data
Addressing AI limitations with enrichment
Inaccurate training data
leads to poor model
accuracy and
performance, yielding
low-quality results
Clean data reduces the
need for extensive data
prep, simplifying the
overall AI pipeline and
improving efficiency
High-integrity data
reduces the time and
computational resources
required for model
development
Practitioners can rely on
consistent data to
extract meaningful
features that contribute
to model performance
Transparent, accurate
data aids in the
understanding of model
decisions, builds trust,
and identifies biases
Data with integrity
avoids introducing noise
that contributes to
overfitting, resulting in
more robust models
Models trained on high-
integrity data are easier
to maintain, as changes
are less likely to cause
unexpected issues
Easier model
maintenance
Reduced
Preprocessin
g Overhead
Effective
Feature
Engineering
Enhanced
Model
Interpretability
Reduced
Overfitting
Faster model
training
Model
Accuracy and
Performance
When AI models are built
on reliable data, they are
more likely to perform
consistently and
dependably
Reliable
Model
Deployment
17 From Noise to Brilliance: Supercharge AI with Data Enrichment
• Financial crimes
and compliance
• Customer insight
• Branch location analytics
• Fraud analytics
• Risk analysis
• Customer insight
• Fraud analytics
• Pricing
• Network and coverage
planning
• Customer insight
• Location-based
marketing & advertising
• Asset management
FINANCIAL SERVICES INSURANCE TELECOMMUNICATIONS
• Customer insight
• Retail location analysis
• Location-based
marketing & advertising
• Home search
• Appraisal analysis
• Valuation modeling
RETAIL
• Service optimization
and delivery
• Planning
• Compliance and safety
• Emergency response
and management
• Economic development
• Site selection
• Market analysis
• Lifestyle modeling
GOVERNMENT REAL ESTATE
• Customer insight
• Checkout analytics
• Logistics and delivery
• Location-based
marketing & advertising
eCOMMERCE
Solve complex, real-world challenges
Key takeaways
Appending relevant context from
additional sources
What is data enrichment?
Accuracy, performance, and utility
across various applications
How does it improve your AI?
Improves business outcomes, saves
money, and user trust
How does it benefit you?
Fireside chat
19
Copyright © 2023 TDWI
QUESTIONS?
CONTACT INFORMATION
If you have further questions or comments:
David Loshin, Knowledge Integrity, Inc.
loshin@knowledge-integrity.com
Antonio Cotroneo, Precisely
antonio.cotroneo@precisely.com
Thanks to Our Sponsor
2
THANK YOU!
Copyright TDWI

More Related Content

PDF
Unlocking the Power of Trusted Data for AI, Analytics, and Business Growth.pdf
PPTX
From Raw Data to Insights: Simplifying Data Validation and Enrichment
PPTX
Top Data Enrichment Companies for Smarter B2B Insights
PPTX
Decision Confidence: Using Modern Approaches to Data Quality to Improve Trust...
PDF
Gabor Koncz – AI in email marketing: email conversion optimization in eCommerce
PPTX
AI You Can Trust - Ensuring Success with Data Integrity Webinar
PPTX
Do You Trust Your Machine Learning Outcomes?
PPTX
(Data) Integrity Matters: Four Ways You Can Build Trust in Your Data
Unlocking the Power of Trusted Data for AI, Analytics, and Business Growth.pdf
From Raw Data to Insights: Simplifying Data Validation and Enrichment
Top Data Enrichment Companies for Smarter B2B Insights
Decision Confidence: Using Modern Approaches to Data Quality to Improve Trust...
Gabor Koncz – AI in email marketing: email conversion optimization in eCommerce
AI You Can Trust - Ensuring Success with Data Integrity Webinar
Do You Trust Your Machine Learning Outcomes?
(Data) Integrity Matters: Four Ways You Can Build Trust in Your Data

Similar to Supercharging AI with Data Enrichment (20)

PDF
Data Integrity for Banking and Financial Services
PDF
Data Integrity for Banking and Financial Services
PPTX
Kickstart a Data Quality Strategy to Build Trust in Data
PDF
Accelerate Confident Decision-Making with Data Enrichment
PPTX
Make more confident business decisions with data you can trust
DOCX
Accelerate Your Career with AI Data Science Certification – Start Now
PDF
Data Innovation Summit: Data Integrity Trends
PPTX
Kickstart a Data Quality Strategy to Build Trust in Data
PPTX
Struggling with Unreliable Data? Learn How to Build Trust in Data for Better ...
DOCX
Get Certified in AI Data Science – Boost Career Growth with This High-Demand ...
PPTX
Kickstart a Data Quality Strategy to Build Trust in Your Data
PPTX
Making Your Data AI Ready: The Critical Role of Data Integration
PDF
-Enrichment - Unlocking the value of data for digital transformation - Big Da...
PDF
Data Integrity Trends
PDF
Data Enrichment using Web APIs
PDF
Crucial Considerations for AI-ready Data.pdf
PDF
Training Taster: Leading the way to become a data-driven organization
PPTX
One Data Governance for Them All – Master Data Included
PDF
Your AI and ML Projects Are Failing – Key Steps to Get Them Back on Track
PPTX
Modern Data Governance:  Synergies with Quality and Observability 
Data Integrity for Banking and Financial Services
Data Integrity for Banking and Financial Services
Kickstart a Data Quality Strategy to Build Trust in Data
Accelerate Confident Decision-Making with Data Enrichment
Make more confident business decisions with data you can trust
Accelerate Your Career with AI Data Science Certification – Start Now
Data Innovation Summit: Data Integrity Trends
Kickstart a Data Quality Strategy to Build Trust in Data
Struggling with Unreliable Data? Learn How to Build Trust in Data for Better ...
Get Certified in AI Data Science – Boost Career Growth with This High-Demand ...
Kickstart a Data Quality Strategy to Build Trust in Your Data
Making Your Data AI Ready: The Critical Role of Data Integration
-Enrichment - Unlocking the value of data for digital transformation - Big Da...
Data Integrity Trends
Data Enrichment using Web APIs
Crucial Considerations for AI-ready Data.pdf
Training Taster: Leading the way to become a data-driven organization
One Data Governance for Them All – Master Data Included
Your AI and ML Projects Are Failing – Key Steps to Get Them Back on Track
Modern Data Governance:  Synergies with Quality and Observability 
Ad

More from Precisely (20)

PDF
What Every Data Leader Should Know About Third-Party Data for AI and Analytic...
PDF
The Future of Automation: AI, APIs, and Cloud Modernization.pdf
PDF
Unlock new opportunities with location data.pdf
PDF
Reimagining Insurance: Connected Data for Confident Decisions.pdf
PDF
Introducing Syncsort™ Storage Management.pdf
PDF
Enable Enterprise-Ready Security on IBM i Systems.pdf
PDF
A Day in the Life of Location Data - Turning Where into How.pdf
PDF
Get More from Fiori Automation - What’s New, What Works, and What’s Next.pdf
PDF
Solving the CIO’s Dilemma: Speed, Scale, and Smarter SAP Modernization.pdf
PDF
Solving the Data Disconnect: Why Success Hinges on Pre-Linked Data.pdf
PDF
Cooking Up Clean Addresses - 3 Ways to Whip Messy Data into Shape.pdf
PDF
Building Confidence in AI & Analytics with High-Integrity Location Data.pdf
PDF
SAP Modernization Strategies for a Successful S/4HANA Journey.pdf
PDF
Precisely Demo Showcase: Powering ServiceNow Discovery with Precisely Ironstr...
PDF
The 2025 Guide on What's Next for Automation.pdf
PDF
Outdated Tech, Invisible Expenses – How Data Silos Undermine Operational Effi...
PDF
Modernización de SAP: Maximizando el Valor de su Migración a SAP S/4HANA.pdf
PDF
Outdated Tech, Invisible Expenses – The Hidden Cost of Disconnected Data Syst...
PDF
Migration vers SAP S/4HANA: Un levier stratégique pour votre transformation d...
PDF
Outdated Tech, Invisible Expenses: The Hidden Cost of Poor Data Integration o...
What Every Data Leader Should Know About Third-Party Data for AI and Analytic...
The Future of Automation: AI, APIs, and Cloud Modernization.pdf
Unlock new opportunities with location data.pdf
Reimagining Insurance: Connected Data for Confident Decisions.pdf
Introducing Syncsort™ Storage Management.pdf
Enable Enterprise-Ready Security on IBM i Systems.pdf
A Day in the Life of Location Data - Turning Where into How.pdf
Get More from Fiori Automation - What’s New, What Works, and What’s Next.pdf
Solving the CIO’s Dilemma: Speed, Scale, and Smarter SAP Modernization.pdf
Solving the Data Disconnect: Why Success Hinges on Pre-Linked Data.pdf
Cooking Up Clean Addresses - 3 Ways to Whip Messy Data into Shape.pdf
Building Confidence in AI & Analytics with High-Integrity Location Data.pdf
SAP Modernization Strategies for a Successful S/4HANA Journey.pdf
Precisely Demo Showcase: Powering ServiceNow Discovery with Precisely Ironstr...
The 2025 Guide on What's Next for Automation.pdf
Outdated Tech, Invisible Expenses – How Data Silos Undermine Operational Effi...
Modernización de SAP: Maximizando el Valor de su Migración a SAP S/4HANA.pdf
Outdated Tech, Invisible Expenses – The Hidden Cost of Disconnected Data Syst...
Migration vers SAP S/4HANA: Un levier stratégique pour votre transformation d...
Outdated Tech, Invisible Expenses: The Hidden Cost of Poor Data Integration o...
Ad

Recently uploaded (20)

PDF
Accuracy of neural networks in brain wave diagnosis of schizophrenia
PPTX
OMC Textile Division Presentation 2021.pptx
PDF
ENT215_Completing-a-large-scale-migration-and-modernization-with-AWS.pdf
PDF
Hybrid model detection and classification of lung cancer
PDF
Mushroom cultivation and it's methods.pdf
PDF
Hindi spoken digit analysis for native and non-native speakers
PPTX
TechTalks-8-2019-Service-Management-ITIL-Refresh-ITIL-4-Framework-Supports-Ou...
PPTX
Tartificialntelligence_presentation.pptx
PDF
gpt5_lecture_notes_comprehensive_20250812015547.pdf
PPTX
A Presentation on Touch Screen Technology
PDF
project resource management chapter-09.pdf
PPTX
KOM of Painting work and Equipment Insulation REV00 update 25-dec.pptx
PDF
Getting Started with Data Integration: FME Form 101
PDF
Web App vs Mobile App What Should You Build First.pdf
PDF
DASA ADMISSION 2024_FirstRound_FirstRank_LastRank.pdf
PDF
Video forgery: An extensive analysis of inter-and intra-frame manipulation al...
PDF
Building Integrated photovoltaic BIPV_UPV.pdf
PDF
Zenith AI: Advanced Artificial Intelligence
PPTX
A Presentation on Artificial Intelligence
PDF
A comparative analysis of optical character recognition models for extracting...
Accuracy of neural networks in brain wave diagnosis of schizophrenia
OMC Textile Division Presentation 2021.pptx
ENT215_Completing-a-large-scale-migration-and-modernization-with-AWS.pdf
Hybrid model detection and classification of lung cancer
Mushroom cultivation and it's methods.pdf
Hindi spoken digit analysis for native and non-native speakers
TechTalks-8-2019-Service-Management-ITIL-Refresh-ITIL-4-Framework-Supports-Ou...
Tartificialntelligence_presentation.pptx
gpt5_lecture_notes_comprehensive_20250812015547.pdf
A Presentation on Touch Screen Technology
project resource management chapter-09.pdf
KOM of Painting work and Equipment Insulation REV00 update 25-dec.pptx
Getting Started with Data Integration: FME Form 101
Web App vs Mobile App What Should You Build First.pdf
DASA ADMISSION 2024_FirstRound_FirstRank_LastRank.pdf
Video forgery: An extensive analysis of inter-and intra-frame manipulation al...
Building Integrated photovoltaic BIPV_UPV.pdf
Zenith AI: Advanced Artificial Intelligence
A Presentation on Artificial Intelligence
A comparative analysis of optical character recognition models for extracting...

Supercharging AI with Data Enrichment

  • 1. A Fireside Chat moderated by David Loshin Affiliate Research Director, TDWI President, Knowledge Integrity, Inc. Senior Lecturer, University of Maryland November 2, 2023 Supercharging AI with Data Enrichment
  • 3. President, Knowledge Integrity, Inc. Senior Lecturer and Lead for External Relations, University of Maryland
  • 4. What we will talk about today • Setting the stage: • Generative AI as a business imperative • Data imperatives for AI model quality • Discussion: The pivotal role of data enrichment in training and fine-tuning Generative AI models
  • 6. What is Generative AI? • A subset of artificial intelligence that includes systems designed to generate outputs such as images, music, text, or other forms of media, based on its training data • Learns from existing and generate new data that is consistent with the original data set • Generative AI systems that have been trained on billions of parameters use prediction to create new instances of data in response to provided prompts • Large Language Models (LLMs) are a type of Generative AI that have been trained on massive amounts of content
  • 7. Ensuring Trustworthy & Appropriate Results Data volumes Data quality Data access • Issues include: – Bias – Privacy – Ethical concerns – Legal concerns – Hallucinations
  • 8. Data Enrichment & LLM Training • Improving the utility of data through appending and integration of relevant content from additional sources • Enrichment is used for – Refining contextual nuances – Improving fidelity of prompt responses – Improve pattern recognition to reduce probability of hallucinations – Improve interpretability of results
  • 10. The leader in data integrity Our software, data enrichment products and strategic services deliver accuracy, consistency, and context in your data, powering confident decisions. of the Fortune 100 99 countries 100 2,500 employees customers 12,000 Brands you trust, trust us Data leaders partner with us 10
  • 11. AI initiatives succeed with trusted data of leading businesses have ongoing investments in artificial intelligence 91% From Noise to Brilliance: Supercharge AI with Data Enrichment Algorithms Data Modeling Large Language Models Deep Learning Hyperparameter Tuning Training Data Retrieval Augmented Generation Supervised Learning Natural Language Processing Bias and Fairness Artificial Intelligence Feature Engineering Neural Networks Chatbots Machine Learning Data Mining 11 Source: NewVantage
  • 12. For trusted data, you need data integrity Data integrity is data with maximum accuracy, consistency, and context for confident business decision-making Data Integrity From Noise to Brilliance: Supercharge AI with Data Enrichment 12
  • 13. What is data enrichment, exactly? 13 It’s the process of enhancing your data by appending relevant context from additional sources – improving its overall value, accuracy, and usability. From Noise to Brilliance: Supercharge AI with Data Enrichment
  • 14. Trusted third-party data at a global scale Addresses & Property Verified and validated address and property data for map display and analytics Boundaries Administrative, community, and industry-specific boundaries for data enrichment and territory analysis Demographics Demographic and consumer context data for better understanding people and behavior Points of Interest Detailed business, leisure, and geographic features for location and competitive intelligence Streets Robust street-level data for mapping, analysis, routing, and geocoding Risk Natural hazard boundaries related to flood, fire, earthquakes, and weather 14 Expertly curated datasets containing thousands of attributes for faster, confident decisions From Noise to Brilliance: Supercharge AI with Data Enrichment
  • 15. 15 From Noise to Brilliance: Supercharge AI with Data Enrichment Purchases & Shopping Building & Parcel Boundaries Lifestyles PreciselyID School Rankings Points of Interest Addresses Population Property Attributes Weather Natural & Manmade Hazards Travel Time Administrative Boundaries Land & Property Consumer Environment Data enrichment can be easy with the right tools A unique identifier for every address that doesn’t change, and other methods for appending data
  • 16. Addressing AI limitations with enrichment Inaccurate training data leads to poor model accuracy and performance, yielding low-quality results Clean data reduces the need for extensive data prep, simplifying the overall AI pipeline and improving efficiency High-integrity data reduces the time and computational resources required for model development Practitioners can rely on consistent data to extract meaningful features that contribute to model performance Transparent, accurate data aids in the understanding of model decisions, builds trust, and identifies biases Data with integrity avoids introducing noise that contributes to overfitting, resulting in more robust models Models trained on high- integrity data are easier to maintain, as changes are less likely to cause unexpected issues Easier model maintenance Reduced Preprocessin g Overhead Effective Feature Engineering Enhanced Model Interpretability Reduced Overfitting Faster model training Model Accuracy and Performance When AI models are built on reliable data, they are more likely to perform consistently and dependably Reliable Model Deployment
  • 17. 17 From Noise to Brilliance: Supercharge AI with Data Enrichment • Financial crimes and compliance • Customer insight • Branch location analytics • Fraud analytics • Risk analysis • Customer insight • Fraud analytics • Pricing • Network and coverage planning • Customer insight • Location-based marketing & advertising • Asset management FINANCIAL SERVICES INSURANCE TELECOMMUNICATIONS • Customer insight • Retail location analysis • Location-based marketing & advertising • Home search • Appraisal analysis • Valuation modeling RETAIL • Service optimization and delivery • Planning • Compliance and safety • Emergency response and management • Economic development • Site selection • Market analysis • Lifestyle modeling GOVERNMENT REAL ESTATE • Customer insight • Checkout analytics • Logistics and delivery • Location-based marketing & advertising eCOMMERCE Solve complex, real-world challenges
  • 18. Key takeaways Appending relevant context from additional sources What is data enrichment? Accuracy, performance, and utility across various applications How does it improve your AI? Improves business outcomes, saves money, and user trust How does it benefit you?
  • 21. CONTACT INFORMATION If you have further questions or comments: David Loshin, Knowledge Integrity, Inc. loshin@knowledge-integrity.com Antonio Cotroneo, Precisely antonio.cotroneo@precisely.com
  • 22. Thanks to Our Sponsor 2