My name is Kay Rottmann and I’m supporting teams in building Artificial Intelligence products and solutions

Driven by a passion for utilizing AI for the improving people’s lives, I bring a diverse range of experience to the table having worked with both startups as member and founder as well as big tech companies. Throughout my career, I have focused on tackling challenging AI problems in areas such as large language models, NLP, computer vision, and reinforcement learning. With a focus on creating impactful solutions, I am dedicated to pushing the boundaries of what is possible with AI and using technology to make a positive difference.

Contact me

Skills and interests

If you ask me about my skills and interests, I would probably answer:

Software Engineering

90%

Research

90%

Leadership

90%

Photography

70%

Diy / Maker

70%

Family

100%

Professional Experience

Stations where I gathered Industry Experience

 
 
 
 
 

Sr. Applied Scientist for Alexa AI

Amazon, LLC.

Jan 2021 – Present
Working on Artificial Intelligence projects for Alexa together with a team of Scientists from multiple locations worldwide focusing on massive multilingual natural language understanding. of Alexa.
 
 
 
 
 

Guest Lecturer

Karlsruhe Institute of Technology (KIT)

Apr 2019 – Present
Giving lectures on Deep Neural Networks and Natural Language Processing for Master students at KIT. Topics involving: “Parallel Computing for Neural Networks and Deep Learning Frameworks”, “Adversarial Attacks”, “Generative Adversarial Networks”, “Explainability”, “InstructGPT/ChatGPT”, “Large Language Models”, “Privacy”
 
 
 
 
 

Head of AI for Manufacturing

Bosch GmbH, Center for Artificial Intelligence

Jan 2019 – Nov 2020 Renningen, Germany

Supporting teams to apply state of the art Artificial Intelligence methods in the manufacturing domain, besides others, working on

  • Automatic Optical Inspection
  • Manufacturing Analytics
  • Scheduling

Previously as Head of Research Engineers supported extraction of artifacts out of state of the art research

 
 
 
 
 

Co-CEO and Co-Founder

strukt.re GmbH

Jan 2017 – Dec 2018 Pfullingen, Germany
Consulting startup to bring artificial intelligence to small and medium businesses. Based on the observation that artificial intelligence is a big success factor in todays world, we provided services to small and medium businesses to make use of this without having their own research engineering team.
 
 
 
 
 

Engineering Manager

Facebook, Inc.

Jan 2014 – Nov 2016 Menlo Park, California, USA

Supported the multilingual text processing team. Our team was responsible for

  • ‘see translation’ scaling up to >1 Billion served translations per day
  • ‘auto translation’, identifying high quality translations and identifying people who need these translations to understand the content
  • scaling up from tens of supported languages to hundreds
  • Language Identification
  • Text classification
  • Multilingual composer allowing people to write posts in multiple languages and providing their own translations
  • Deep learning based translation
 
 
 
 
 

Technical Lead

Facebook, Inc.

Sep 2013 – Jan 2014 Menlo Park, California, USA
Building and launching of first in-house machine translation engine for translation of user posts in Facebook.
 
 
 
 
 

Research Scientist

Mobile Technologies GmbH / Jibbigo LLC

Oct 2007 – Sep 2013 Karlsruhe Germany
Built first completely on-device speech to speech translator for iOS as promoted by Apple: https://www.youtube.com/watch?v=oNyTWBrAr68

Academic Background

Education and academic activities

Neural Networks Guest Lecturer

Giving lectures on Deep Neural Networks and Natural Language Processing for Master students at KIT, Topics covered include

  • Deep Learning
  • Image Recognition / Segmentation / Classification
  • Generative Adversarial Networks
  • Adversarial Attacks
  • Large Language Models
  • InstructGPT/ChatGPT
  • Privacy

PhD in Computer Science (Artificial Intelligence)

“Online Incremental Machine Translation” Allowing automatic learning of improvements to speech to speech translation systems from user feedback across multiple users.

Built a distributed architecture that instantaneously updates and incorporates user feedback to improve translation quality without tedious training cycles.

Computer Science Diploma (comparable with Master)

Best Computer Science Diploma in 2007, winner of Klaus-Tschira Award Major: Computer Science, Artificial intelligence, Minor: Mathematics

Scholarship for writing Diploma Thesis at Carnegie Mellon University

Advisors Prof. Dr. Alexander Waibel and Dr. Stephan Vogel, Carnegie Mellon University, School of Computer Science

Recent Posts

Just some things I tried out

After two posts on linked in by Jim Kring about hard working cats, I became curious on some other biases that exist in image generation and I started to play around with DALL-E. I can’t say how much is true for other providers like Midjourney but I assume similar results will hold true. I started playing around with it mainly interested in effects in a multilingual setup, however I found a few other things I want to point out below.

  1. ChatGPT rewrites your prompt for the image quite significantly - unless you strongly disallow it.
  2. Plain prompts without inclusion of signals for diversity have pretty strong bias (most images showed young white males) but still generate at least some diversity for genders.
  3. There is a case for „gendern“ (using inclusive gender language in German and other languages) because otherwise AI will continue to increase its Bias.
  4. Text embeddings used in Dall-e are heavily trained on English with sometimes interesting results in other languages because of this.

Looking back, I realize my previous article didn’t delve deeply into the potential impact of recent AI advancements. While I can’t predict the future with certainty, I’m ready to make informed predictions about how current AI developments may disrupt traditional systems. In a nutshell, I foresee a significant, imminent impact across various industries. While IT, Finance, and Marketing are often touted as the prime sectors for AI disruption, I believe Manufacturing also stands at the cusp of substantial transformation. With the power of AI, smaller enterprises might disrupt the markets currently dominated by big players, potentially leading us into a scenario reminiscent of the next Innovator’s Dilemma. Furthermore, I anticipate an urgent need for individuals to adapt to these changes, with the integration of new AI tools into educational curriculums being an ideal approach.

Ah, large language models (LLMs), GPT-4, and generative models in general… It seems like everybody’s talking about them these days. So, naturally, I felt the unbearable pressure of groupthink to also write about it. As someone who’s been in the NLP field for quite some time, I have seen the transformative power of these models firsthand. In this post, we’ll explore the potential of LLMs as enablers and force multipliers, transforming what individuals can achieve in various fields. Just imagine the possibilities!

In 2022 we exchanged the heating in our house including solar thermal energy as well. Since I like to know what is going on in the house I was interested to read out and store the information from the heating. However as it turns out the manufacturer doesn’t allow this unless you pay for a subscription to send the data to their servers and then access it from there. Without the subscription all that was possible was to use an app to read the current state of the system. This got me thinking whether I could reverse engineer the app communication to extract the data I need without a subscription.

This is part of a lecture I gave at KIT (Karlsruhe Institute of Technology) in the master course “neural networks”. It is to show what adversarial attacks are and what they mean for machine learning based classification.

Contact

Get in Touch