Abdulkader
Artificial Intelligence Engineer & Data Scientist based in Vienna.
I grew up in Aleppo, Syria. I studied informatics at IMC Krems and did my masters in AI at JKU Linz. I'm currently based in Vienna, Austria. I arrived in Austria as a refugee in 2015, and I worked my way from unskilled jobs such as a security guard, warehouse worker, to data science and AI. My focus has been on efficient fine-tuning techniques, memory optimization, and building data-driven systems that leverage LLMs that reduce workload. Beyond AI, I'm interested in finance, and passionate about philosophy. I'm also a 19th century history nerd. To wind down, I often play strategy games.
I taught myself to code during night shifts as a security guard, completed internships at Microsoft and STRABAG, and recently finished my Master's focusing on deep learning and applied AI.
Selected Impact:
- Light for the World: Built a highly specialized AI harness for an agent to automate the generation (of parts) of the strategy evaluation, reducing workload from months to days, slashing external consulting costs. Built a system to automate data workflows which otherwise performed manually, reduced workload from days to minutes.
- Baubot: Developed robotic data-collection suites for model training which successfully solved a critical problem in the robot positioning. Improved client-facing 3D simulation software by reworking algorithm which boosted performance significantly, and implemented features with direct contact with clients.
Iām currently looking for full-time Machine Learning Engineer / Data Scientist roles in a company that fits my values where I build high-impact AI systems with the team.
Current topics, research areas, and questions I am exploring:
Parameter-efficient fine-tuning methods have excessively proliferated, but it's one of the most creative subdomains of deep learning, and remains immensely valuable in the industry. I keep myself informed on the latest research.
I made my philosophical conclusions a long time ago and I'm at peace with them, but the questions on consciousness are forever pervasive to the human discourse. Lately, there has been some research on AI consciousness which I find provocative and intriguing.
I'm always down to talk about finance, markets, and portfolio management.
My favorite subdomain in AI. Among all paradigms, I find Reinforcement learning the most profound. It's the most similar to how we learn as humans which is why I tend to romanticize it a bit.
Six books I've read and recommend with no specific order:
Often called the bible of reinforcement learning. It's a dense book written by the pioneers of the field. Finishing it felt like a great milestone and in my opinion, reinforcement learning in general has some of the most beautiful mathematics in the field. I keep going back to it and I highly recommend it to everyone in the field.