This keynote takes the audience on a journey into the future and the year 2029, where computers have as high processing power as the human brain. This means machines can do pretty much everything a human can do, only faster and better. What will the world look like?
This is not a doomsday talk, on the contrary. Perhaps we´ll then have no poverty, famine or environmental issues. This lecture tells how this can happen.
How WWF is using oil and gas data from Rystad Energy to find the hidden facts about Norwegian oil and gas industry - and how they use it to influence the political discourse.
#0 Plenum
10:00 - 10:30
Pause
10:30 - 11:00
Digital transformasjon i bygg og anlegg. Fra hammer og spiker til kunstig intelligens
Har du noen gang opplevd at datainvesteringene kunne ha gitt enda mer verdi? Hvordan kan vi sikre at dataplattformer blir tatt i bruk på best mulig måte? Vi mener svaret på dette er god dataadopsjon!
Gjennom flere år har Capgemini hjulpet Widerøe med å utvikle og ta i bruk en ny dataplattform. Hør hvordan dette har gjort Widerøe mer datadrevet og hvordan det har økt omsetningen med 20%.
#2 Data
10:30 - 10:55
Modern Data Architecture on AWS | Read more
By Kate Johnsen Solutions Architect, Amazon Web Services
A modern data architecture acknowledges the idea that taking a one-size-fits-all approach to analytics eventually leads to compromises. It is not simply about integrating a data lake with a data warehouse, but rather about integrating a data lake, a data warehouse, and purpose-built stores, enabling unified governance and easy data movement. With a modern data architecture on AWS, customers can rapidly build scalable data lakes, use a broad and deep collection of purpose-built data services, ensure compliance via a unified data access, security, and governance, scale their systems at a low cost without compromising performance, and easily share data across organizational boundaries, allowing them to make decisions with speed and agility at scale.
#4 Data Architecture
10:30 - 10:55
Developing uncertainty-aware neural networks with limited amounts of data | Read more
Training neural networks (NNs) for production can be challenging, especially when data is limited. In these circumstances we want our model to reflect the fact that it cannot (or should not) be very confident about its predictions, as data tends to drift with time and thus lie outside the training data support. Sadly, research has shown that NNs are over-confident when applied to out-of-distribution data; this behavior is unintuitive and unwanted in applications where NN output is used as critical decision support.
In this talk I will discuss how to regularize neural networks and make them more aware of the context they were trained in: I will discuss how to teach your NN to tell if incoming data is out of distribution, and at the same time yield better calibrated output. These and related methods will all be discussed in the context of a concrete business case that I recently worked on.
Digital transformation has influenced changes in the data & analytics industry over the last several years, leading to the creation of a modern data stack. In this session, you will learn about the history of a modern data stack, how the data world evolved to necessitate it, and what a modern data stack looks like today.
We will discuss:
• How strides in technology enabled the current state of data platforms
• The essential components & features that make up a modern data stack
• What data teams need to know to take advantage of emerging trends
#4 Data Architecture
11:00 - 11:25
Scaling AI across the organization with MLOps | Read more
By Xiaopeng Li AI Business Lead, Western Europe at Microsoft
Scaling AI is challenging. It takes people, platforms and processes to effectively move AI solutions from PoC into production. More and more companies have adopted machine learning operations (MLOps) to streamline machine learning lifecycle management and scale AI solutions across their entire organizations.
In this talk, Xiaopeng will explain the why, what and how of MLOps with practical guidance. You will also hear how organizations like FedEx and SAS are leveraging MLOps today to accelerate AI-infused innovation and deliver tangible business outcomes.
#3 Smart
11:00 - 11:25
En introduksjon til geografisk informasjon og Geonorge | Les mer
En relativt rask gjennomgang av hva geografisk informasjon er, hvor man finner det og hva det kan brukes til.
Geonorge er en del av Norge digitalt; et samarbeid mellom offentlige virksomheter med ansvar for å etablere og forvalte kartdata og annen stedfestet informasjon.
#2 Data
11:00 - 11:25
The principles used to create value from data at Oda (and scaling from 3 to 75 data practitioners in three years) | Read more
Oda is a Norwegian grocery scale up (backed by Softbank, Prosus, Kinnevik ++) now going international and launching in Finland and Germany this year. Data has been a key differentiator succeeding with online groceries, both as core parts of the retail system (e.g. inventory management, route planning, product recommendations) and operational model (e.g. forecasting, campaign planning, insights ++).
Leser man om data governance på LinkedIn eller følger nyheter fra konsulenthusene så sitter man igjen med en oppfatning om at data governance handler om å kjøpe en datakatalog, bry seg om "data lineage" og datakvalitet og på magisk vis så skal dette gjøre deg "datadrevet". Slik er det selvfølgelig ikke, men det er vanskelig å selge software som skal endre en hel kultur i en virksomhet.
Data governance – som annen styring – handler om mennesker, organisasjon og kultur rundt data, mer enn selve dataene. Data governance er ikke svaret, men prosessen for å komme opp med de riktige svarene. I Fremtind Forsikring driver vi disse prosessene, vi er datadrevet, men vi er langt fra kommet i mål. Vi vil dele den innsikten vi har opparbeidet oss så langt.
Abstract Norway and Europe are said to be lagging behind in the digital race, especially when it comes to development and application of AI driven solutions. But, even if Europe is perhaps not known for its native tech giants, we have several strongholds that can make it easier to benefit from the technological breakthroughs in more inclusive ways.
So, how can we use our unique innovation system to develop a viable IT sector that create new jobs, a more inclusive society and green transformation? What collaborative initiatives do we need to look out for? This will be the focus of Marianne’s talk.
#1 Make
12:00 - 12:25
DNB's journey towards “data-as-a-product” in a distributed data architecture | Read more
A socio-technical paradigm shift is sweeping through large enterprises such as DNB where data sharing and governance is radically changed through Product-Thinking.
We discuss in this workshop how to differentiate the unique perspectives that are involved in designing data products that are to be shared across domains in an organisation.
#4 Data Architecture
12:00 - 12:25
Anyone can tune a model, but can you make a model worth tuning? | Read more
Because ML research focuses on models, it is easy for industry practitioners to fall into the trap and do the same. The use of Kaggle as a platform for learning ML amplifies this problem. Ruter’s machine learning team spends less than 5% of their time on model development.
Where do they spend their remaining time? They are scoping projects, identifying use cases, and generating value from existing models. Because of this, Ruter has flipped the deck and switched from developing machine learning models based on performance metrics to developing models based on usability.
Here we will present the evolution of our load prediction model. It predicts the load for ~500 000 daily departures across ~200 routes for busses, trams, and boats with more than ten different types of sensors.
Hvordan bekjempe kriminalitet ved hjelp av data? Hva betyr informasjon og analyse i arbeidet med å forebygge, forhindre og oppklare saker? Og hvilke muligheter ligger i automatisert informasjonsbehandling i politiet?
Politiets IT-enhet er politiets samarbeidspartner på IT-siden. Vi er en egen organisasjon med over 600 engasjerte nerder som utvikler og videreutvikler smarte løsninger for politiet. Vi vet at effektivt politiarbeid ikke kan gjøres uten fremtidsrettet teknologi og verdfull data. Politiets aktiviteter skal være kunnskapsbaserte - derfor er innsikt ett av våre absolutte satsingsområder.
Potensialet er stort og mulighetene er mange. I dette foredraget får du innblikk i et datadrevet politi - nåtid og framtid.
#3 Smart
12:30 - 12:55
Fremtidens informasjonsforvaltning i Helseetaten til Oslo kommune | Les mer
Hvordan koordinere 18 selvstendige virksomheter som skal utføre en tjeneste med regelverk som forandrer seg stadig?
Hvordan gå fra 25 silo-baserte fagsystemer til ett selvbetjent datanettverk og fremme en datakultur hvor tjenesten driver utvikling?
Vi forteller historien om hvordan Helseetaten planlegger å bli en forvalter av data rundt innbyggeren, i en modell som støtter samhandling mellom forskjellige virksomheter og fagområder.
TINE har de siste årene vært i en prosess fra å skifte organisasjonens innsiktsløsninger fra tradisjonelt on-prem datavarehus og BI-verktøy over i en ny og moderne arkitektur i skyen. I denne sesjonen forteller vi om erfaringer vi har gjort oss i denne prosessen samt hvorfor vi valgte Data Lakehouse til å drive TINEs både tradisjonelle rapporteringsbehov og avansert analyse.
#4 Data Architecture
13:00 - 14:00
Pause
14:00 - 14:25
Datadrevet prosessforbedring gjennom process mining | Les mer
Hvordan overvåke og forbedre prosesser ved hjelp av process mining?
Hvordan kan en leder forstå hvordan prosesser utføres i praksis, og hvor det er best å justere for å forbedre organisasjonens måloppnåelse?
I UDI har vi tatt i bruk process mining og vi forteller gjerne om vår reise mot å bruke data for å utvikle bedre prosesser.
#1 Make
14:00 - 14:25
Hvordan en krevende true crime tvang fram et nytt verktøy for sikker og oversiktlig gravejournalistikk | Les mer
I prosessen med å lage true crime-serien «Vålnes-saken» (november, 2021) holdt NRKs journalister på å drukne i en en komplisert etterforskning med 1300 dokumenter, 400 vitner og en svært lang tidslinje.
Løsningen ble et spesialutviklet researchsystem som funket så bra at det nå skal piloteres for hele NRK. På veien har det blitt til et trygt sted for dokumenter og analyse, med en standardisert metadatasett for undersøkende journalistikk. Det brukes også eksterne maskinlærings-APIer for berikelse av datasettet.
I denne presentasjonen får vi bli med bak sceneteppet for å lære mer om hvordan NRK lagde dokumentarserien, hvordan de jobber med data og fortsatt tar vare på kildevern, hvordan dette systemet beriker metadata med maskinlæringsapier og hvorfor det av og til lønner seg å ha en hacker med på laget.
#3 Smart
14:00 - 14:25
Driving digital transformation with a world class data platform | Read more
Mesta is the biggest road maintenance and operation service provider in Norway. It starts its digitalization journey by designing and implementing a world class and future proof data platform in Azure. We manage to continuously deliver high quality insights that engage the whole organization to be more analytical and data driven with very limited number of resources.
#4 Data Architecture
14:00 - 14:25
How to intelligently accelerate the journey to become a data-driven organisation and boost business outcomes? Read more
How do you make agile data-driven decisions a reality across an organisation from being an ever-growing aspiration? How can you boost productivity and business outcomes to delight your business stakeholders?
This session works through the most significant challenges organisations face when striving for data-driven expansion and how to address them. Considerations around addressing technical debt, data literacy within the workforce, collaboration on data and leveraging AI by fuelling it with Metadata will be addressed:
Understanding your data footprint and considerations around technical debt, what does simplification and standardisation really mean and look like in practise?
Data at your fingertips, what is really needed to empower your organization with the right data with the right quality, to the right person, at the right time?
Driving outcomes and customer value by leveraging a modern cloud data management platform.
#2 Data
14:30 - 14:55
Grønne data for bynatur og innbyggerinvolvering | Les mer
Hvordan brukes data i arbeidet med å sikre Oslos bynatur? Og hvordan kan vi involvere innbyggerne som deltakere i planleggingen, fremfor å være passive mottakere av informasjon? Andreas vil vise eksempler og løfte ideer og visjoner.
#2 Data
14:30 - 14:55
Smart bruk av data innen avdekking av forsikringssvindel | Les mer
Forsikringssvindel påfører forsikringsbransjen flere hundre millioner i tap hvert år i Norge, og fører til økte priser for alle oss med forsikring. Tidligere har identifikasjon av svik i forsikringsoppgjør vært en manuell prosess basert på enkle regler, heuristikk og tidvis også magefølelser, men Storebrand har i de siste årene hatt en satsning på å finne smartere metoder for å avdekke forsikringssvindel.
I 2021 kom Storebrand til finalen i Innsiktsprisen basert på vårt arbeid innen bruk av maskinlæring og avansert analyse for avdekking av forsikringssvindel. Denne presentasjonen vil gå litt mer i dybden på det tekniske arbeidet til Storebrand som ble presentert for Innsiktsprisen, samt snakke litt om det som har skjedd i Storebrands arbeid innen avdekking av forsikringssvindel siden.
#3 Smart
14:30 - 14:55
REMA 1000s reise mot avansert analyse: Et konkret data science-eksempel for pristilbud | Les mer
Beskrivelse av reisen for å ta data science inhouse i REMA 1000 og utnytte potensialet til avansert analyse av store datamengder. Arkitektur+system+mennesker.
Konkret eksempel på hvordan REMA 1000 beregner salgseffekt av pristilbud for å ta bedre beslutninger om hvilke produkter, hvilke prispunkt som er best og bedre styring av verdikjeden tilknyttet pristilbud ved bruk av maskinlæring.
#1 Make
14:30 - 14:55
AI architecture - strategies & roadmap for a systems engineering’s approach | Read more
Based on the guest lecture I will be doing this fall with MIT, I would like to present real world cases and frameworks for AI strategies with implementation roadmaps.
AI is revolutionizing many industries, including energy, consumer products and services, automotive, financial services, national security, healthcare, and advertising. But too often, business and IT leaders take a limited view of AI, focusing almost exclusively on machine learning (ML) methods. But AI technologies are, in fact, key enablers to complex systems. They require not only ML technologies, but also trustworthy data sensors and sources, appropriate data conditioning processes, responsible governance frameworks, and a balance between human and machine interactions. In short, organizations must evolve into a systems engineering mindset to optimize their AI investments.
In this talk I will give an overview of how to get started, from strategy, to framework and practical tools needed for successful implementation.
#4 Data Architecture
15:00 - 15:30
The past, present, and future of DNB’s self-service Insight Platform | Read more
Over three years ago, DNB presented its vision for a “Data Kingdom” at the Make Data Smart event. Since then, it has been quite a ride for DNB to bring this vision to fruition. In this session, Olav will share DNB's experiences about what has been achieved and where the organisation is headed. A significant aspect of the platform that DNB has built over these years is its self-service nature. As and when new business use cases are to be onboarded, the business users can cherry-pick the platform capabilities they need and automatically provision them. It conforms perfectly to the “self-serve data infrastructure as a platform to enable domain autonomy", a principle of Data Mesh, about which DNB will talk in detail in a dedicated session.
Attend this session to learn more about DNB’s Insight Platform and how 150+ data scientists and analysts use it.
#4 Data Architecture
15:00 - 15:25
How a 375 year old company tries to be more data-driven | Posten's utilizations of Data Science | Read more
How does one go forward as a 375 year old company trying to become more data-driven?
Posten holds a rich tradition of being one of the largest logistics operators in Norway, having delivered letters and parcels to its citizens for centuries.
In this talk, i try to paint a picture of how Posten is slowly but steadily becoming more data-oriented in its solutions, and how the data science department has evolved since its conception in 2019. The aim is to offer a transparent case of not only what's gone well, but what lessons the company has learned and is continually learning as the journey unfolds.
#1 Make
15:00 - 15:25
Building an Investment Simulator in NBIM | Les mer
Hver måned utbetaler Statens pensjonskasse pensjon til 340 000 nordmenn. I 2021 utgjorde dette totalt 31 milliarder kroner.
I deler av saksbehandlingen er reglene så kompliserte at hver sak ble kontrollert av to saksbehandlere. Ved hjelp av datadrevet innsikt og maskinlæring har vi utviklet modeller som har erstattet saksbehandler nummer to gjennom både å identifisere feil – og til å avdekke feil før de oppstår.
I dette foredraget får du lære mer om hvordan vi bruker maskinlæring til å forutsi pensjonsutbetalingene, sikre kvalitet og effektivisere arbeidsprosesser.
#3 Smart
15:25 - 15:45
Pause
15:45 - 16:15
How Data-On-Demand Architectures Lead to Green Architectures | Read more
Organizations store more and more data in ever-larger volumes. However, most of that data is not new or original, but copied. Companies excel at duplicating data. For example, information about a customer is stored in a CRM system, a staging area, a data warehouse, several data marts, and a data lake. Even within one database, data is stored multiple times to support different users. In addition, copies of data are stored in development and test environments. There is also data redundancy between organizations when exchanging data. Usually, the receiving organization stores the data in its own systems, resulting in even more copies.
This unrestrained duplication of data has many disadvantages and challenges, including higher data latency, complex data synchronization solutions, more complex data security and privacy enforcement, higher development and maintenance costs, higher technology costs, and more complex database and metadata administration. Additionally, all this data copying involves processing and storage which costs energy.
In this session, Rick van der Lans explains how you can design data-on-demand architectures in which data copies and unnecessary processing are minimized. The technology is available for developing such architectures. The extra benefit is that data-on-demand architectures are also greener architectures.
#0 Plenum
16:15 - 16:40
Paneldebatt: Data, AI and Analytics in 3 and 10 years m/Xiaopeng Li, Lars Rinnan, Nina Walberg og Marianne Jansson Bjerkman | Les mer
Årets tema for paneldebatten er:
- Data, AI og Analytics om 3 og 10 år
- - - Strategier, kompetanse og teknologier
- GDPR, datasikkerhet og dataflyt
--- Hvordan sikre både dataflyt og samarbeid, samt sikker data behandling
- Etisk AI
- - - Hvordan sikrer vi at AI blir brukt for gode formål, og sikre (utilsiktet) misbruk
Deltakere:
- Lars Rinnan - CEO Amesto Nextbridge
- Xiaopeng Li - AI Business Lead Microsoft
- Nina Walberg Head of Data & Insight, Oda
- Marianne Jansson Bjerkman - Cluster Manager AI Smart Innovation Norway
#0 Plenum
16:45 - 18:00
Mingling m/drikke og snacks i utstillingsområde | Les mer
Etter konferanseslutt møtes vi i utstillingsområde for mingling og noe godt å drikke og snacks. Det blir også konkurranser og premieutdeling.
#0 Plenum
* Programmet oppdateres fortløpende og endringer vil forekomme.