Applied AI Lunch Talk
AI will drive audio into the next era
When humans interact socially via speech, our brains do an extraordinary job converting pressure waves into neural encodings of phonemes, words, sentences and concepts that we can make sense of, reflect on, and respond to. Some times, the challenge is too hard, like if you are in a noisy party and head for the dance floor because it just feels too tiring to have a conversation.
Augmented hearing is setting out to bring the ideal sound scene to you at all times. Just like when your smartphone camera enables you to take beautiful photos by blurring the background, making the subject even more clear and vivid. This is enabled via AI. For good reasons, AI tools are rarely applied to your audio scenes. Bridging AI with neuro science and audio insights, we envision a future where people can have clear and fluent conversations without being unnecessarily drained from poor audio conditions.
Bio: Mette Carstensen
Graduated from DTU as M.Sc. Biomedical Engineering in 2017. Specialised in ML, DSP and Auditory Signal processing, generally focused on signals from and functional images of the brain. Worked with Digital Transformation for McKinsey & Co, and subsequently with developing audiological algorithms for hearing aids at Demant. Genuinely passionate about bringing technology to improve people’s lives.
How can AI/ML help to make best use of the worlds agricultural resources?
by FOSS
At FOSS our mission is to provide data and insights to help the food industry make optimal use of raw materials and reduce food waste. In this presentation we will share insight on the journey we are on to apply data science and machine learning to address some of the big challenges the world is facing. It is a journey full of possibilities, but also a journey which emphasizes a range of practical challenges that needs to be addressed to apply data science to support a truly data driven food production system.
BIO: Toke Lund-Hansen
He has a background in physics and computer science and holds a Ph.D. from the Technical University of Denmark. He has been working for FOSS during two periods initially from 2011 developing technology for X-ray and spectroscopy instruments. After a few years in a start-up, he now manages a team of specialist developing our spectroscopic, x-ray, and vision technologies and their integration in FOSS’ products. Lately, with a strong focus on vision and AI.
BIO: VP for frontend Innovation, Thomas Nikolajsen
Thomas holds a Ph.D in Optical Science from DTU. Since then he has worked with innovation in industry and various startups with focus on how to apply new technology to address key problems associated with sustainable food production. He has been with FOSS since 2020 heading a cross discipline innovation team responsible for scoping the development of the company's future solution.
Personalized treatment for neurodegenerative disorders through applied deep learning approach in combination with advanced wearables
Parkinson’s is the second largest neurodegenerative disorder and is the fastest growing one. The current golden standard for diagnosing movement-related disorders like Parkinson’s is unfortunately still based on a completely subjective standard scale. This leads to several major consequences for the overall society and especially in the long term with a 47% misdiagnose rate in primary care and often results in a very late diagnosis. We at Paragit Solutions combine the latest hardware wearable technology with applied deep learning to analyze and cluster potential upcoming patients based on their abnormal muscle fluctuations during the day.
BIO: Mohammad Filfil
He is the founder, and CEO of Paragit Solutions and is leading the overall product development. Filfil received a master’s degree in electronic engineering from DTU in 2019 and San Jose State University -Silicon Valley. He is the inventor of two patents used in Paragit’s hardware sleeve. Previously co-founded 3 other startups within deep tech and hard tech and have mostly been the lead developer of hardware products.
Bridging the epilepsy treatment gap through an AI enabled, low cost diagnostic solution
by BrainCapture
80% of people with epilepsy live in low- and middle-income countries where EEG services are hard to access. BrainCaptures aims to bridge the epilepsy treatment gap through an AI enabled, low cost diagnostic solution. This presentation provides an insight into BrainCaptures' solution.
BIO: Tue Lehn-Schiøler
He is the CEO of BrainCapture. He holds a PhD in machine learning and has been applying AI to real world problems for more than 15 years.
AI Due diligence
by Amesto
One remarkably interesting case involved a due diligence study of data and machine learning aspects of a start-up that a venture capital fund was considering investing in. We will describe how we approached this systematically and the focus areas of our study, as well as our experience in this area from our AILab initiative.
BIO: Laura Frølich
She is Lead Data Scientist at Amesto NextBridge Denmark. Previously, Laura applied data science in the water sector and supervised MSc students at DHI, did consulting at Teradata and researched non-specific effects of vaccines at the Danish SSI. Laura holds a PhD in machine learning from DTU.
NLP on Incident reporting
by Amesto
During the COVID crisis, we applied Natural Language Processing techniques to aid the International Federation of Red Cross and Red Crescent automate their categorization of disaster response reports globally.
BIO: Emilie Lundblad
She is Managing Director at Amesto NextBridge. Former quant & startup entrepreneur, turned Data & AI nerd. Cand.Oecon from SDU, ML & AI from MIT, MSc. IT Management from ITU - writing thesis on AI. Guest speaker at ITU & MIT 2022, Make Data Smart, Genesis, Dansk IT.
Transforming an old labour-based service to an automated machine solution based on AI
Non-Destructive Testing (NDT) for large critical structures, like wind turbines.
Development has gone a slow transformation based on the need of the industry. When the need to go places where we can send humans, like nuclear reactors, we solved it by automating the inspection and data collection. Automated data collection was a success, and the last 30 years data collection has been standard part of our services.
On the wave of new disrupting companies in the 2010s, our CTO asked the question: How could we disrupt our self? Big data, cloud storage and automated data evaluation was on the list. Since we have been working on improving the total inspection speed with rapid high-quality data and fast algorithm-based evaluation.
BIO: Mikael Fuglsang
He is project manager in Digital Service at FORCE Technology, changing the way our customers work with data from their quality inspection. Working in the field of building automated mechanics solution that’s tailored to fit their product inspection and training AI models to become specialise in analysing the data for quality control.
Design the future
How do we design a fantastic future? What are the practical applications of AI and how do we make it happen? Johan will talk about building AI products in the real world, the future we’re headed towards, and invite the audience for an open dialog.
BIO: Johan Jessen
He has spent the last 10 years in Silicon Valley designing advanced technology in fields such as AI, robotics, biotech, aerospace, and self-driving cars. Previously, he led the Autonomous Design team at Lyft and launched moonshots at Google X.
Can AI help create the perfect cup of coffee
by BKI
Is it possible to use AI to predict how a cup of coffee will taste? BKI believes so. Nicholaj will be taking you through the journey BKI has been through building an AI product that can help create the right blend of coffee beans to create the required taste of a cup of coffee.
BIO: Nicholaj Månsson Olsen
Nicholaj has spent the last 7 years at BKI building and implementing systems to digitize their production and create the foundation for advanced analytics. He has led projects related to data warehousing, digitization, and automation of processes and AI/Machine learning.
Human-centric AI
by Ambolt
In his talk Andreas will take you for a deep dive of how Applied AI can generate value today. He presents how Ambolt AI helps domain experts boost their efficiency with AI. Ambolt AI calls this Human-centric AI.
BIO: Andreas Berre Eriksen
Co-founder of Ambolt AI has a background in Computer Science, and is on a mission to empower people with AI. At Ambolt AI we don’t believe AI will replace people, but we do believe people, who use AI will replace people who don’t.
How can we solve the increasing need for ‘Quality AI’?
by Amira Moussa from Cheetah AI
The demand for faster, more robust, and more energy-efficient AI has increased exponentially in the last few years while the development cycle has been stuck. Can the FPGA chip solve this issue?
AI dilemma: A Pilot or Co-pilot
by Sara Shafiee & Alireza Taghizadeh from DivERSNowadays, most aspects of our lives have been substantially affected by AI. With the widespread use and huge power of AI, we can ask an important question: Should AI be the pilot or co-pilot? In this talk, Sara and Alireza will talk about how DivERS uses AI to help recruiters to remove conscious and unconscious biases from the recruitment process. Above all, they will explain how DivERS brings humans into the loop, by letting us be in the front seat and AI as a co-pilot.
Improving Patient Outcomes with AI: How Corti is Changing the Game
by Corti
In healthcare, the ultimate goal is to improve patient outcomes. In this talk, we’ll examine how Corti’s innovative technology is transforming the way healthcare providers assess and treat patients. We’ll explore real-world case studies from Corti as well as the future of AI technologies in healthcare.
Bio: Henrik Cullen
Henrik Cullen is the VP of Product at Corti AI. He has led the development of most of the product and enjoys working closely with customers to improve the experience for healthcare professionals and patients. Henrik is committed to continuing to develop innovative AI solutions that make a real difference in the lives of patients and healthcare providers.
Navigating the AI Act: A Collaborative Approach Between Engineers and Lawyers
by UCPH
The forthcoming AI Act challenges existing legislative paradigms, emphasizing self-criticism and multidisciplinary collaboration. This talk explores those new compliance requirements for multidisciplinarity, and looks at how we can create better collaboration models between lawyers and engineers.
BIO: Léonard Van Rompaey
Postdoc at the Law faculty - Centre for Private Governance at University of CopenhagenWith a focus on law and AI since 2014, my award-winning research has led to collaborations with organizations such as Teknologisk Instituttet and the Danish robotics sector, and appearances in Danish media. While I primarily teach at Copenhagen's Law Faculty, I've also led workshops for law firms, consultancies, and public entities.
Using AI to Create Value From the Data Haystack
by Novozymes
This talk will introduce three Novozymes applied AI use-cases, as well as the motivations behind, executions and considerations taken. We will also discuss how the recent surge in LLMs (e.g., ChatGPT) has affected the company, and what factors to recognize when incorporating such models in the day-to-day of business.
BIO: Victoria Payne & Peter Mørch Groth
Victoria is a Data Engineer in Advanced Analytics at Novozymes. She works with sourcing and processing data for data-driven projects as well as maintaining compliance and establishing guidelines for the use of AI in Novozymes. Victoria values keeping up to date with new technologies while still possessing a critical outlook on which of these can truly surpass the hype and create value.
Peter graduated from DTU as M.Sc. in Mathematical Modelling and Computation in early 2022. With a specialization in machine learning, he is now an Industrial PhD student at Novozymes and the University of Copenhagen. His research is focused on machine learning-driven enzyme engineering.
Utilizing AI/ML in large manufacturing processes
by ROCKWOOL
As we navigate the fourth industrial revolution, the integration of artificial intelligence and machine learning has revolutionized how we optimize production, enhance quality control, and streamline operations. Join us for a talk where Shabab, Senior Data Scientist at ROCKWOOL A/S, will discuss and explore the challenges and opportunities surrounding the implementation of AI and ML solutions, shedding light on the ease and complexities involved in reshaping the future of manufacturing.
Bio: Shabab Akhter, Senior Data Scientist at ROCKWOOL
Shabab has a background in Computer Science and over 5 years of experience working at the intersection of technology and business. He combines his deep technical expertise and ability to easily and effectively communicate complex technical terms to deliver end-to-end solutions. In his career so far, he has been responsible for both managing and developing AI/ML solutions that transformed business processes in multiple industries.
The Power of the Personal: Making AI for Individuals
by LAIKA
LAIKA's founders believe that the next stage of AI will focus on the personal: crafting unique experiences for each individual user. They've been exploring this for the last year with a personalised creativity tool for writers, and are now set to pivot to crafting a full-fledged "writers room" with AI. In this talk, they'll share wisdom from their journey so far, including 3 specific actions startups can take to personalise their AI offerings.
BIO: Charlene Putney, COO, Write With LAIKA
Charlene Putney is an award-winning writer and teacher from Ireland. After working at Google and Facebook in management positions, she's been writing for videogames since 2013, including writing for Divinity: Original Sin 2 and Baldur's Gate 3.
BIO: Martin Pichlmair, CEO, Write With LAIKA
Dr Martin Pichlmair is Associate Professor at the Creative AI Lab in ITU, and a veteran entrepreneur with several games studios under his belt including the multi-award-winning Broken Rules.
Harnessing computer vision and overcoming the data challenge
by Focalx
In this talk, we'll dive into how focalx harnesses traditional AI, specifically computer vision, as a foundational pillar of its business model. We'll explore the challenges posed by data scarcity in training ML models and unveil the innovative solutions we've adopted. Discover how, by employing generative AI and other cutting-edge techniques, focalx creates synthetic data to bridge the gap, ensuring more accurate and reliable vehicle inspections.
BIO: Isabella Merrild Agdestein, CEO & Co-founder at focalx.ai
Isabella, once a pro dancer, has since delved into the world of Autotech entrepreneurship. Her experiences span Strategy & Implementation Consultancy and are further enriched by her Master studies in Finance & Strategy. She founded focalx, a B2B SaaS startup disrupting vehicle inspection and handovers using AI.
Physical Informed Neural Network for hybrid condition monitoring
At FORCE Technology we are investigation how to improve hybrid testing, by using machine learning as an approach to address the complex challenges in condition monitoring. We are integrating advanced neural network techniques with an understanding of the underlying physical principles, this cutting-edge methodology offers a powerful solution to efficiently analyze and predict the behavior of assets.
Bio: Mikael Fuglsang, Team Lead for Robotics and Project Manager Digital Services at Force Technologies
Changing the way our customers work with data from their quality inspection. Working in the field of building automated mechanics solutions tailored to fit their product inspection and training AI models to become specialized in analyzing the data for quality control.Introducing the new AI Test and Experimentation facility in DK
EU is investing heavily in addressing the gap between European and international capabilities in AI. One aspect is the ability to test AI before entry to market and DTU is coordinating the one for Smart Cities and Communities AI and robotics domain, known as CitCom.ai.
The purpose of CitCom is to accelerate new AI and Robotics innovations to the market and will offer support especially for SMEs and their cutting-edge solutions and provide test that can serve to get the first difficult customer.
Bio: Rasmus Sune Reeh, Program Leader at the European Testing and Experimentation Facility
Rasmus Reeh is project lead of CitCom.ai and deputy at Connecting Communities at DTU Compute Dynsys. An economist by training he has worked more than 10 years within different roles in the Smart City domain, including DTU Elektro and latest as Development Lead of Copenhagen Solutions Lab.
The Talk:
How generative AI can drive business transformation and successThis talk will provide insight into how Implement helps clients with challenges that can be solved through generative AI. From high-level strategic initiatives that assess a business’s readiness to apply and adopt generative AI, to the development of proof-of-concept through co-creation, and lastly, the deployment of these solutions to create impact.
Bio: Victor Emil Funch, Data Science Consultant at Implement
Victor Emil Funch is a Data Scientist at Implement and has been working with data science, machine learning, and specifically natural language processing since 2019. Upon finishing his master’s in economics in 2022, he joined the Implement Data & Analytics department which is part of the Digital Transformation practice.The Talk:
Creating Systematic Reviews in a fraction of the timeJose will delve into the transformative power of AI, showcasing its ability to revolutionize the creation of "Systematic Reviews" in a fraction of the time compared to traditional procedures. Through real-world applications, he will share valuable insights on how Erudita.tech utilizes AI to achieve Scientific Impact, Enhanced Productivity, and CO2 Reduction, generating significant value. Join us for a compelling discussion that unveils the immense potential of AI and explores the challenges encountered on this transformative journey with Erudita.tech, Jose's innovative entrepreneurial project.
Bio: Jose Cerdán, CEO of Fysiolutions JC ApS
With a background in healthcare and IT, a PhD in telerehabilitation, and a postdoctoral position in telemedicine, Jose brings over 24 years of experience at the intersection of technology and health. His expertise lies in "Health Digitalization and Research," and he is passionate about AI and how this technology can be applied to optimise "Scientific Text Production, Access to Scientific Knowledge and Science Communication". Jose has successfully founded four startups in health and technology, securing over €7M in soft money for R&D projects from EU and private funds.
The Talk:
How can dementia be detected in speech with AI? The DemensAI approachWith over 55 million people affected worldwide and ranking as the 7th leading cause of death, dementia represents a significant health challenge. However, the healthcare system is under enormous pressure and is looking for better tools and innovative solutions to streamline diagnosis without adding to the workload.
Bio: Laurine Dargaud, CTO of DemensAI
Laurine holds a DTU MSc degree in Human-Centered Artificial Intelligence, and she developed the proof of concept for DemensAI: a speech-based model that can detect Alzheimer's disease based on a voice recording of picture description.Bio: Abhista Partal, CSO of DemensAI
Abhista holds a DTU MSc degree in Autonomous Systems, with a specialization in Explainable AI. Abhista is dedicated to AI research and development at DemensAI, experimenting with the latest speech and language models to improve dementia detection.The Talk:
AI for Mental HealthThis talk presents research in AI for emotion recognition and behavioral coding and how it can be used in mental health. The applications in health include research in psychiatry, an emotionally aware AI coach, and couple therapy. The talk will also discuss ethical concerns like biases in AI.
Bio: Line Clemmensen, Co-founder, CTO of Tetatet and Soil, Associate Professor at DTU Compute
Line is a Professor at DTU Compute, co-founder and CSO of Tetatet, and co-founding Soil. Her research is focused on machine learning and artificial intelligence with applications in psychiatry, amongst other things.