This document discusses how data and AI on Azure can help higher education institutions in several key areas:
1) Achieving a consistent hybrid learning experience for students on and off campus through conversational AI and digital learning assistants.
2) Simplifying end-to-end research processes by addressing big data and AI requirements through Azure services like Machine Learning and Cognitive Services.
3) Assessing student digital interactions and results to predict potential wellbeing issues using analytics on Azure.
4) Building an end-to-end Azure Modern Data Platform to drive business insights from various institutional data domains.
1. Data and AI in Higher Education
Ignacio Borrego
Azure Data and AI Lead
UK Education
2. Agenda
• Value of Data – Key Themes
• Key Problems and Solutions
• Modern Data Platform – how can it help?
• Synapse – what is it, why should I care?
• How have we helped? Case Studies
• How do Learning Assistants fit in?
• Research requirements – Big Data and AI
• Q&A
3. The Value of Data and AI on Azure
Adapting to a Hybrid Learning Model
How can we achieve a consistency of experience?
Achieve an engaged and individual learning
experience both on and off student campus, through
the use of conversational AI and DLAs on Azure.
Research Requirements – Big Data and AI
Simplifying the end-to-end research process
Addressing researchers’ Big Data and AI
requirements on Azure ensures an efficient, cost-
optimized and secure environment through services
like Azure ML, Cognitive Services, or HDInsight.
Student Engagement and Wellbeing
Detecting risks in the ‘new normal’
Assess students’ digital interactions, results and
engagement to predict any potential wellbeing
issues before they become a serious problem
through analytics on Azure.
University Business Planning
Data Discovery and Actions for Improvement
From student, to estates, to HR data, build an
end-to-end Azure Modern Data Platform
to drive useful business insights from a wide
range of different data domains.
4. Different Questions – A Single Platform
"We want to monitor
student wellbeing,
proactively detecting
at risk students."
"We want to forecast
our national and
international student
numbers and
associated revenue."
“We want to plan the
resources we will
need for the coming
academic year.”
5. The Problem
“Our data is siloed
within our institution,
held in different
sources, hard to
consolidate.”
“Our data comes in a
large variety of formats
– structured,
unstructured, semi-
structured…”
“We have data in
different places: on-
premises, VLE, Azure,
other cloud
providers...”
Modern Data PlatformAt the core of all use cases is…
6. What is a Modern Data Platform?
A Modern Data Platform is a future-proof architecture for Business Analytics. It is a functional
architecture which has all components to support
- Modern data warehousing
- Machine Learning and AI development
- Real-time data ingesting & processing.
10. Experimentation
Fast exploration
Semi-structured data
Big Data
OR
Businesses are forced to maintain
two critical, yet independent analytics systems
Proven security & privacy
Dependable performance
Operational data
Relational Data
Data Lake Data Warehouse
11. Integrated analytics platform for AI, BI, and continuous intelligence
Platform
Azure
Data Lake Storage
Common Data Model
Enterprise Security
Optimized for Analytics
Data lake integrated and Common Data
Model aware
METASTORE
SECURITY
MANAGEMENT
MONITORING
Integrated platform services
for, management, security, monitoring,
and metastore
DATA INTEGRATION
Analytics Runtimes
Integrated analytics runtimes available
provisioned and serverless
Synapse SQL offering T-SQL for batch,
streaming, and interactive processing
Synapse Spark for big data processing
with Python, Scala, R and .NET
PROVISIONED (DW) SERVERLESS
Form Factors
SQL
Languages
Python .NET Java Scala R
Multiple languages suited to different
analytics workloads
Experience Synapse Studio
SaaS developer experiences for code
free and code first
Artificial Intelligence / Machine Learning / Internet of Things
Intelligent Apps / Business Intelligence
Designed for analytics workloads at any
scale
Azure Synapse Analytics
16. “We wanted to provide a common view.
An Executive Dean of a faculty should be
able to see the number of students they
have, the amount of space they have,
and the amount of research income
they’re able to access—all in a single
place.”
Matt Gordon: Associate
Director of Analytics
19. “The improved transparency has been
extremely beneficial. It has led to a
much deeper shared understanding of
our organization and greater
consensus around what we need to
prioritize moving forward. Evidence-
based decision-making has enabled
the organization to achieve greater
levels of equity and innovation.”
Richard Salter: Director of
Analytics
22. Andrew Proctor – PVC Digital
“Going to university can be stressful and is often the first
time a teenager will move out …Beacon is there to help
them, it’s not just a Q&A bot. In ‘Welcome Week’ it can
recommend societies, [helping] them make friends, and
will eventually guide them towards services if they need
more support … It will ask them how their lectures are
going to ensure that if they are struggling, we can help
them more quickly and in the best way.”
23. How did they achieve this?
• Partnered with ANS – Microsoft Expert Partner
• Solution built with a fast time to market on the
Azure AI Stack
24. Intelligent Agent → Key Source
Agent on Azure -> Student Data Signals
-> Monitoring Wellbeing
29. Domain specific pretrained models
To simplify solution development
Azure
Databricks
Machine
Learning VMs
Popular frameworks
To build advanced deep learning solutions
TensorFlowPyTorch ONNX
Azure Machine
Learning
LanguageSpeech
…
SearchVision
Productive services
To empower data science and development teams
Powerful infrastructure
To accelerate deep learning
Scikit-Learn
Familiar Data Science tools
To simplify model development
CPU GPU FPGA
From the Intelligent Cloud to the Intelligent Edge
Azure Notebooks JupyterVisual Studio Code Command line
30. Jason Atkin, Assistant Professor
University of Nottingham
- “One of the things cloud computing
does is bring the power and data
processing ability of huge machines to
any researcher’s desk.”
31. Andrew Blake
Research Director, The Alan
Turing Institute
“Working with Microsoft, our growing community of
researchers have been tooled up with skills and access to
Azure for cloud computing and as a result they’ve been
able to undertake complex data science tasks at speed.
We look forward to growing our engagement with the
Azure platform to help us to undertake even bigger and
more ambitious research over the coming academic year.”