image-description
8-10 April
Cinemaxx, Copenhagen

Schedule.
Get an overview of all the sessions.

cloud cloud
Looking for Slides from the talks?

Expand a session entry in the schedule to see the abstract and links to slides and other related resources.

09:00 10:00
Room 1 (IMAX)
From Power BI to Power AI –
Enrich your Business Intelligence
with Artificial Intelligence
Markus Ehrenmüller-Jensen
  • Data Platform
  • Everyday AI
  • Business Intelligence

Download Slides from this session - demo files available here

AI is everywhere – and now even included in Power BI. Sometimes AI might be very apparent when you enrich your data with predictions by explicitly calling an Azure Machine Learning web service in Power Query. Sometimes it might by hidden in a nice little context menu when Power BI is automagically explaining the reason of a difference or a change over time in your data to you.
No matter if you are a business user, analyst or data scientist – Power BI has AI capabilities tailored to you. In this session we will cover how you can integrate and leverage the use of language R, how to integrate an Azure Machine Learning Service when loading data, what kind of insights Power BI is capable of delivering automatically, how you can create a complex new column without typing a single formula/function, how you can explore your data by asking questions in plain English, ... and more!
Sounds like a lot of content? I omitted most of the slides in favour to spend the time in live demos instead. Come in this session to learn how you can enrich your very own Business Intelligence with Artificial Intelligence and make the step up from Power BI to Power AI.

Level: Advanced

Markus Ehrenmüller-Jensen Savory Data
BI Architect
Room 6
Secrets of the Cosmos Marilag Dimatulac
  • Data Platform

Azure CosmosDB is a very powerful cloud database. With its promise of high-availability, low-latency, global-distribution, the possibilities are as vast as the universe. But how can we unlock its secrets without burning our resources like a supernova? Let’s take CosmosDB out for a spin and explore how to design our data architecture, tune performance, and optimize cost so we can truly build intelligent, planet-scale applications

Level: Intermediate

Marilag Dimatulac Dewise
Solutions Architect
Room 7
Differentiate Big Data vs Data Warehouse
use cases for a cloud solution
James Serra
  • Data Platform

Download Slides from this session

It can be quite challenging keeping up with the frequent updates to the Microsoft products and understanding all their use cases and how all the products fit together. In this session we will differentiate the use cases for each of the Microsoft services, explaining and demonstrating what is good and what isn’t, in order for you to position, design and deliver the proper adoption use cases for each with your customers. We will cover a wide range of products such as Databricks, SQL Data Warehouse, HDInsight, Azure Data Lake Analytics, Azure Data Lake Store, Blob storage, and AAS as well as high-level concepts such as when to use a data lake. We will also review the most common reference architectures (“patterns”) witnessed in customer adoption.

Level: Intermediate

James Serra Microsoft
Big Data/Data Warehouse Evangelist
Room 8
Start your datacentre transformation journey
with Azure Migrate
Sarah Lean
  • Cloud Management

Download Slides from this session

Have you ever wondered what your environment would look like in the Cloud? How much would it cost? What would be the migration path? With Azure Migrate you and start to answer those questions. The tool can be run in your environment to help understand the tight interdependencies between applications and workloads that exist, then give you an analysis of what your Azure costs would look like. 

From this session you will walk away with a good idea of how to start utilising Azure Migrate in your organisation.

Level: Intermediate

Sarah Lean Microsoft
Cloud Solution Architect
Room 9
How to Get Customers up and Running
When your Data Center is Hit by Lightning
Kennie Pontoppidan
  • Customer Story
  • DevOps
  • Cloud Management

Slides for this session will not be made available

Using the real-world story of our response to a recent weather-induced regional Azure outage, I will show you the Monitoring, Diagnostics, and Analytics tools and data we use to run Dynamics 365 Business Central (and all other Azure services). 

From the initial monitoring alerts through to the final post mortem analysis, I will take you through the incident and the entire data lifecycle to reveal how we maintain control and perspective on the service and the KPIs we use to drive data-driven decision making, even when lightning strikes!

Think of this session as a factory floor tour inside incident management of Azure.

Level: Intermediate

Kennie Pontoppidan Microsoft
Program Manager
Room 10
Predicting earthquake damages with
Azure Machine Learning
Eva Pardi
  • Everyday AI

Blog post with content from the Session (as the session contained mostly demos, no slides are available)

In April, 2015 there was an earthquake with 7.8 magnitude, and an epicenter in the Ghorka District of Nepal. The disaster injured more than 30000 people, which was in most cases caused by the collapsed buildings in the earthquake. 

During the demo-oriented session you receive hands-on knowledge to build your own predictive Azure Machine Learning model, to shape it for your own business and to improve it for better results. 

With this model you can investigate the risks of damage in case of another disaster. The data we use for the demo can be used to mitigate which buildings might need strengthening, so not only many buildings, but also thousands of lives could be saved in another earthquake.

Level: Advanced

Eva Pardi Laerdal Copenhagen
Data Scientist & Software Developer
10:00 10:20
Coffee Break
10:20 11:20
Room 1 (IMAX)
Enterprise BI with Power BI
and Analysis Services Part 1
Christian Wade
  • Business Intelligence

Slides not yet available

Power BI and Analysis Services enables you to build comprehensive, enterprise-scale analytic solutions. This session will deep dive into exciting new and upcoming features. Various topics will be covered such as management of large, complex models, connectivity, programmability, performance, scalability, management of artifacts, source-control integration, and monitoring. Learn how to use Power BI Premium to create semantic models that are reused throughout large, enterprise organizations.

Level: Advanced

Christian Wade Microsoft
Senior Program Manager
Room 6
Using Cloud to spur innovation Rasmus Hald
  • Customer Story
  • Cloud in the Enterprise

Slides not yet available

Cloud is in many ways a paradigm shift, and if we do not change the way we work, we will not get the full value out of a move to Public Cloud Computing. This story comes straight from the trenches from our work at MAERSK, and will give you a unique insight into how we are changing the game with the move to Public Cloud. Learn the good, the bad and the ugly from our ongoing journey to the cloud.

Level: Intermediate

Rasmus Hald Maersk
Head of Cloud Center of Excellence
Room 7
Getting started with
Python Machine learning in SQL Server
Terry McCann
  • Everyday AI
  • Data Platform
  • DevOps

Download Slides from this session

Everyone needs to start learning machine learning, by 2020 80% of all applications will be powered by a form Artificial Intelligence (Machine learning - don’t worry the robots are not rising). Machine learning is no longer just for data scientists, everyone working with data need to have a basic awareness of machine learning. If you work with data, you should be investing in machine learning. 

In this session we will look at Python as a language and explore its packages for interactive machine learning. Terms like SkLearn, Pandas, SciPy, Pickle will become familiar to you by the end of this session. You won't be an expert in machine learning but you will know how to get started with Python. This session we will build models in Python for SQL Server, then we will deploy them. 

Level: Intermediate

Terry McCann Advancing Analytics
Microsoft MVP & Data Science Consultant
Room 8
From poll to push. Azure Event Grid deep dive Sam Vanhoutte
  • Cloud in the Enterprise
  • IoT & Streaming Data

Download Slides from this session

Event driven architectures are difficult to build. Distributed event handling is quite complex. And quite often, a lot of compute power is consumed in polling for events that are not there. Azure Event Grid heavily simplifies building event driven, serverless microservices architectures.

Azure Event Grid is publishing a lot of Azure related events that your application and services can react to. In this session, Sam will give an overview of Azure Event Grid and mention several real life scenarios where Event Grid could be used.
You will learn how to react to events, by using event handlers. You will also see how to publish custom events from your application and the concepts of Event Topics and Event Subscriptions will also be explained.
Please expect a lot of demos.

Level: Advanced

Sam Vanhoutte Codit
CTO, focus on Cloud & Innovation
Room 9
Azure B2C: Application Security Made Easy Sjoukje Zaal
  • Identity & Security

Slides not yet available

Join me to secure your custom applications with Azure Business-to-Consumer, a Cloud identity service on Azure which provides out of the box identity providers like Facebook, Microsoft Accounts, Google+, LinkedIn, and many others, or you can add your own and leverage these in your applications as well.

We will look at the basics of Azure B2C, we will set up an application using Azure B2C for authentication, use built-in policies, and much more.

Level: Advanced

Sjoukje Zaal Ordina NL
Principal Expert Microsoft | Azure MVP
Room 10
Data Science: Predictive Modeling with R Dejan Sarka
  • Business Intelligence
  • Everyday AI

Download Slides and demo files from this session

Many people thing first about predictions when they hear the term data science. And of course, predictive modeling is one of the most important data science tasks. There are many supervised, or directed algorithms for this task. In this session, you will learn about the most important algorithms like neural networks and logistic regression. You will be exposed to the knowledge about all kind of trees and forests, including decision and regression trees, random forests, and gradient boosting trees. A simple algorithm K-nearest neighbors can be useful for some occasions as well.

Level: Advanced

Dejan Sarka SolidQ
MVP, MCT, NTNK
11:20 11:40
Break
11:40 12:40
Room 1 (IMAX)
Enterprise BI with Power BI
and Analysis Services Part 2
Christian Wade
  • Business Intelligence

Slides not yet available

Power BI and Analysis Services enables you to build comprehensive, enterprise-scale analytic solutions. This session will deep dive into exciting new and upcoming features. Various topics will be covered such as management of large, complex models, connectivity, programmability, performance, scalability, management of artifacts, source-control integration, and monitoring. Learn how to use Power BI Premium to create semantic models that are reused throughout large, enterprise organizations.

Level: Advanced

Christian Wade Microsoft
Senior Program Manager
Room 6
Adding AI to your SQL Server applications with
Containerized Cognitive Services
Anna Thomas
  • Everyday AI

No slides available, but notebook with code/demos from the session can downloaded via github

You don't have to be a data scientist to build data applications with AI! Learn how to leverage AI via Cognitive Services in containers in SQL Server (on-prem and Azure) to develop intelligent data solutions.

After this session, you should:

- Understand how using SQL Server ML Services + Extensibility has become easier for ML model development and leveraging prebuilt AI 

- Understand how to implement Cognitive Services in Containers with SQL using the Java Extensibility feature

- Be able to leverage the resources provided to call containerized Cognitive Services from SQL Server in your new and existing data applications

Level: Intermediate

Anna Thomas Microsoft
Data & Applied Scientist
Room 7
Azure Data Warehouse: Why, What and How Nico Jacobs
  • Data Platform

Slides not yet available

If your regular SQL Server becomes too slow for running your data warehouse queries, or uploading the new data takes too long, you might benefit from the Azure Data Warehouse. Via its “divide and conquer” approach it provides significant performance improvements, yet most client applications can connect to it as if it is a regular SQL Server.
But to benefit from these performance improvements we need to implement our Azure Data Warehouse in the right way. In this session - through a lot of demos - you will learn how to setup your Azure Data Warehouse (ADW), review indexing in the context of ADW and see that monitoring is done slightly different from what you’re used to.

Level: Intermediate

Nico Jacobs U2U
SQL Server, BI and AI trainer
Room 8
Serverless real use cases and best practices Thiago Almeida
  • Serverless
  • Customer Story

Download Slides from this session

Let's take a look at real cases from Microsoft customers worldwide who have solved their problems through Azure's serverless & messaging services. We'll cover their projects, architecture decisions, and best practices or lessons learned from developing these solutions!

Level: Advanced

Thiago Almeida Microsoft
Software Engineering Lead
Room 9
Lessons from the API Management trenches Eldert Grootenboer
  • Cloud Native
  • Cloud in the Enterprise

Download Slides and Postman Collection via Github

Azure API Management has a lot of functionality, but it's not always clear when to use what. In this session we will go into setting up an API Management architecture, inspired by real life use cases. We will see how we can expose and protect our services, which policies help make our life easier, and how to handle our application lifecycle management.

Level: Advanced

Eldert Grootenboer Motion10
Cloud Solution Architect - Azure MVP
Room 10
Innovative pharma in the cloud Christian Holm Diget
  • Customer Story
  • Cloud in the Enterprise
  • Cloud Native

Slides not yet available

Innovation has been held back in highly regulated industries for a long long time. Actually, just delivering working software in a reasonable time frame has been elusive. Requirements to comply with GxP, HIPAA, part 11 and various other standards and pieces of legislation makes the development process difficult and requires loads of documentation. 

However things are changing, even in the most regulated companies. In this talk we will dive into how Trifork eHealth are partnering with Microsoft to help companies like Lundbeck become innovative using Azure PaaS to build a controlled cloud native setup and deliver software faster.

Level: Intermediate

Christian Holm Diget Trifork eHealth
TechLead
12:40 13:40
Lunch
13:40 14:40
Room 1 (IMAX)
Take control of your BI projects with Azure DevOps David L. Bojsen
  • DevOps
  • Business Intelligence
  • Data Platform

Slides not yet available

In this session we will demonstrate a fully working BI agile development environment from the end user’s side – business decision makers who will make demands and ensure the quality, developers providing the functionality and architects and project managers making sure that quality, budget, timelines etc. are as desired, while delivering in an agile and mostly automated delivery method. To explore the “geeky” stuff on how to setup everything – visit our other session “Automate your BI Project with Azure DevOps”

Level: Intermediate

David L. Bojsen Kapacity
Principal Architect
Room 6
Azure Security Center unleashed Tom Janetscheck
  • Identity & Security
  • Cloud Management

Download Slides from this session

In cloud environments, management is increasingly distributed, attackers continue to innovate, and thus, cloud security management looks like mission impossible.

Join this session for a deep-dive into Azure Security Center, witness on-stage live attacks against an Azure environment and learn what you need to know in order to secure an Azure environment.

Level: Expert

Tom Janetscheck Alegri
Microsoft Azure MVP | Cloud native
Room 7
Driving Power BI automation through monitoring Jan Mulkens
  • DevOps
  • Business Intelligence

Slides not yet available

You finally got the go ahead and now you have a nice and shiny Power BI service or perhaps a Power BI Report Server environment.
Maybe you went all out and even have set up a deployment pipeline to automate deployments.

The one thing you probably haven't don't yet, is setting up that feedback loop.
You're missing metrics. The important metrics that enable you to manage your environment beyond 10 users.

So many things to think about...
Who's using your reports, how often and at what times?
Who actually needs a pro license, who doesn't need it anymore?
You need to plan for maintenance, you need impact assessments for outages or deployment failures.
And you surely need these metrics to show the validity of Power BI within your department or even the enterprise.

In this session you'll learn everything you'd ever want to know about monitoring your Power BI environments and how you too can start monitoring Power BI Report Server or the Power BI service like a pro!

Level: Intermediate

Jan Mulkens Ordina BE
Microsoft Data Platform & BI Consultant
Room 8
The way of the exploding workflow -
Become a Logic App Expert
Steef-Jan Wiggers
  • Serverless
  • Cloud Native

Download Slides from this session

“Do you want to be an awesome low-code developer?” You will become the expert, like a karateka reaching the 10th Dan after this session. Over the years you as a developer might have faced many versions of workflow like K2, SharePoint, or Windows Workflow Foundation (WF). Furthermore, you might look for one in the cloud. Well, there’s some good news - the latest addition to the workflow in Azure is Logic Apps, a mature service enabling you to build business flows using a low-code approach. However, low-code you can still leverage your developer skills by integrating with Functions, leverage expressions, fiddle around in the code-behind file of a workflow definition, and build ARM templates. Moreover, before you know it, you will many of them - performing all kinds of tasks, and integrations to support your business processes. In this session, we will go on a journey to reach the 10th DAN using the analogy of the epic Commodore 64 game “the way of the exploding fist”!

Level: Advanced

Steef-Jan Wiggers CoditNL
Senior Consultant, Azure, Integration, SeaHawks Fan
Room 9
Microsoft CNTK Leila Etaati
  • Everyday AI

Slides not yet available

In this session, an overview of what is deep learning and how it works will be presented. Next, an overview of a different type of algorithms available for deep learning will be explained. Then, how we able to use them for the aim of prediction, image processing and so forth will be discussed. Finally, an overview of CNTK, the Microsoft deep learning tool, will be provided and the audience will learn more about it.

Level: Intermediate

Leila Etaati RADACAD
Microsoft MVP
Room 10
Azure PowerShell vs Azure CLI: Duel at the command line Aleksandar Nikolic
  • Cloud Management
  • Cloud in the Enterprise

Slides not yet available

For a couple of years, Microsoft has had two command-line tools for managing Azure resources. Azure PowerShell and Azure CLI. Both tools are cross-platform and open source. Both are frequently updated and support the fast pace of the Azure changes. You can use them locally or in the Azure Cloud Shell. They provide a similar user experience; however, they are not exactly the same. Some differences are easily visible, some differences happen behind the scenes. What is the right tool for you? Do you need to pick one, or are there any benefits in using them together? Join us to learn about their capabilities, and pros and cons of using them, to help you make a choice.

Level: Intermediate

Aleksandar Nikolic PowerShell Magazine
Microsoft Azure & CDM MVP
14:40 15:00
Ice Cream Break
15:00 16:00
Room 1 (IMAX)
Advanced Data Factory: Let the Data Flow! Simon Whiteley
  • Data Platform
  • Business Intelligence
  • Cloud in the Enterprise

Slides and code/demos available via github repository

Modern Azure Data Factories are lean, efficient data pipelines, fully parameterised, dynamic and hugely scalable - this session will show you how to achieve this and more!

Many people still miss the "good old days" of SSIS, and it's now possible to run SSIS within Data Factory, but that doesn't resolve the scalability and automation problems it faces. But never fear, new ADF Data Flows provide a slick, GUI interface over an incredibly scalable, cloud-first compute backend.

This session will run through the new features in ADFV2 and discuss how they can be used to streamline your factories, putting them in the context of real-world solutions. We will also look at the new compute options provided by ADF Data Flows, review how it interacts with Azure Databricks and set you up for truly cloud-native ETL!

Level: Intermediate

Simon Whiteley Adatis
Data nerd, cloud herder, markitect & geek
Room 6
8 lessons learned running K8S
with Azure Kubernetes Service
Pascal Naber
  • Cloud Management
  • Cloud in the Enterprise
  • Cloud Native

Download Slides from this session

Learn how to run Kubernetes cloud native on Azure.

In this session, Pascal shows you how to use Kubernetes in practice using a mix of Azure services. He demos best-practices and the most common scenarios to run containerized .NET Core applications on AKS. This includes using the Key Vault, Storage, Monitoring, how to apply DevOps, .NET Core tips and much more.

Join this session to find out what it takes to run your Docker containers on Kubernetes using Azure.

Level: Advanced

Pascal Naber Xpirit
Azure architect
Room 7
Aggregations in Power BI Marco Russo
  • Business Intelligence

Slides not yet available

Aggregations have been introduced in 2018 in Power BI, as an optimization technique to manage large tables. By providing pre-aggregated tables, you can highly improve the performance of a Tabular data model.
In this session we introduce the concept of aggregation, we show several examples of their usage understanding the advantages and the limitations of aggregations, with the goal of building a solid understanding on how and when to use the feature in data models.

Level: Expert

Marco Russo SqlBI
SQLBI
Room 8
DevOps for Artificial Intelligence,
the road to production
Henk Boelman
  • Everyday AI

Slides and code available via blogpost

With machine learning becoming more and more an engineering problem the need to track, work together and easily deploy ML experiments with integrated CI/CD tooling is becoming more relevant than ever.

In this session we take a deep-dive into the DevOps process that comes with Azure Machine Learning service, a cloud service that you can use to track as you build, train, deploy and manage models. We zoom into how the data science process can be made traceable and deploy the model with Azure DevOps to a Kubernetes cluster.

At the end of this session you have a good grasp of the technological building blocks of Azure machine learning services and can bring a machine learning project safely into production. 

Level: Intermediate

Henk Boelman Heroes
Microsoft AI MVP
Room 9
A guide through the Azure Messaging services Eldert Grootenboer
  • Cloud in the Enterprise

Download Slides from this session

Almost any solution we build will have some sort of messaging in their architecture. But how do we choose what type of messaging we need? From messages to events, from streams to publish and subscribe, the options seem overwhelming. In this session we will dive into the various types of messaging, and which services we can leverage in Azure to implement them. Once finished, you the world of messaging is at your feet, and you will be able to counter any type of messaging which comes on your path.

Level: Advanced

Eldert Grootenboer Motion10
Cloud Solution Architect - Azure MVP
Room 10
Azure IoT Hub for client data collection Rasmus Wätjen
  • IoT & Streaming Data
  • Serverless

Slides not yet available

In this talk, I will show how to use Azure IoT Hub not for telemetry or two-way communication, but how to use it for data collection on a bigger scale.
Rich client software, that has to deliver blobs of data, e.g. for machine learning of larger datasets than telemetry, can also use IoT Hub. 
Using Azure IoT Hub Device Provisioning with X.509 (TLS) certificates. I will walk through how to create a certificate authority and a chain of trust, so clients can authenticate with the IoT Hub Device Provisioning Service without having been pre-provisioned in a device factory. 
Then I will show how to create a client in .NET which can contact the provisioning service and become attached to the IoT Hub and send data.
I will also demo a serverless app that receives the data from IoT Hub and moves the data for further processing later.

Level: Advanced

Rasmus Wätjen Widex
Azure wrangler at Widex
16:00 16:20
Refreshment Break
16:20 17:20
Room 1 (IMAX)
ETL with Azure Databricks, The Complete Introduction André Kamman
  • Data Platform
  • Cloud in the Enterprise

Slides not yet available

Azure Databricks seems to be the new sheriff in town. It promises easy but capable ETL (amongst other things). But how does it work? For an SSIS (and possibly BIML) person this all looks new and strange. Python notebooks? Scala? Spark? What is all this, and where do I start? Well, you start by coming to this presentation. We’ll take a look at how to approach making an Azure Databricks based ETL solution from start to finish. Along the way it will become clear how Azure Databricks works and we will use our SSIS based ETL knowledge to see if it can handle the common use-cases from our daily jobs. And of course we will also have some fun and see how fast it can go!

Level: Intermediate

André Kamman Data Masterminds
Data Platform Architect
Room 6
Azure Automation in production,
lessons learned in the field
Andreas Sobczyk
  • Cloud Management
  • Cloud in the Enterprise

Slides not yet available

Take control of your Azure Automation environment(s). All the tips and tricks we have learned in the field and from our own production environment to help you get a full 360 cycle for your automation. This session will guide you through from runbook design/template using helper modules through easier ways to manage your or your customer's runbooks. Manage your runbooks using PowerShell or in combination with source control to setup a fully featured continuous delivery pipeline. Finally, it will show how to do save up to 2 year of logs for a low cost using Log Analytics and how to report using PowerBI for a great overview. Automate the Automation!

Level: Advanced

Andreas Sobczyk CT Global
Senior Consultant
Room 7
Regain control with Azure Governance Sam Cogan
  • Cloud in the Enterprise

Download Slides from this session

If you've been using Azure for any length of time, the question of how you govern and control the resources you and your team create will have come up.

Azure now offers a number of tools that you can leverage to start to understand what you have deployed, what it is costing you and then to start regaining control over these resources. In this session, we will look at what these tools are, how they work and how you can implement them. Starting with some simple audits we'll work all the way up to strict policies and resource definitions. We'll cover Policies, Blueprints, Resource Graph and Cost Management, individually and working together to provide a complete governance solution.

Level: Intermediate

Sam Cogan Willis Towers Watson
Cloud Solutions Architect
Room 8
How to get the most from Azure SQL Database Vitor Pombeiro

Slides not yet available

In this session I will show how to get the most advantages out from Azure SQL Database.
We will learn how to monitor the performance of our workload, how to improve it and how to secure it using the tools and features at our disposal on the Azure portal without having to install 3rd party tools.
We will also see some ways to lower the costs or increase performance without having to also increase costs.

Level: Intermediate

Vitor Pombeiro Microsoft
Support Engineer for Azure SQL
Room 9
Securing an Azure Function REST API
with Azure Active Directory
Rick van den Bosch
  • Identity & Security

Download Slides from this session - Source code on github, and resulting website

With Azure Functions it’s pretty simple to implement a REST API. With Azure Active Directory it’s not too hard to secure an application. Angular enables you to develop rich front-ends. But how do you go about putting these three together…?

This session will show you how to secure your Azure Functions powered REST API using Azure Active Directory, and how to consume this API from an Angular front-end.

Level: Advanced

Rick van den Bosch Betabit
Cloud Solutions Architect
Room 10
Persona based chit-chat in conversational bots Anshuman Suri
  • Everyday AI

Slides not yet available

Most often, chat-bots are built to solve the purpose of a search engine or a human assistant: their primary goal is to provide information to the user or help them complete a task. However, these chat-bots are incapable of responding to unscripted queries like “Hi, what’s up”, “What’s your favorite
food”. Even if they can, their tone often sounds scripted and lacks personality. Human evaluation judgments show that humans come to a consensus on the intent of a given query which is from chat domain only 77% of the time, thus making it evident how non-trivial this task is.

In our session, we will show why it is difficult to break the chitchat space into clearly defined intents: a system to handle this task in chatbots, keeping in mind scalability, interpretability, appropriateness, trustworthiness, relevance and coverage. We introduce a pipeline for query understanding in chitchat using hierarchical intents as well as a way to use seq-seq auto-generation models in professional bots. We show how various components such as adult/offensive classification, grammars/regex patterns, curated personality based responses, generic guided evasive responses and response generation models can be combined in a scalable way to solve this problem. In addition to these features, we will showcase work which helps personalize the responses of bots and the possibility of injecting dynamic content into auto-generated responses.

Level: Advanced

Anshuman Suri Microsoft
Applied Scientist
image-description

Ready to save your seat?

Buy ticket

Conference Partners

Microsoft Denmark Kapacity unit it