About
My journey into AI was not a conventional one. It began with a deep curiosity about the human mind, leading me to pursue degrees in Clinical and Organizational Psychology. This foundation gave me a unique lens through which to view artificial intelligence—not just as a set of algorithms, but as a reflection of the cognitive processes I had studied. In 2019, I fully committed to this path, earning a Master's in AI (Cum Laude) and diving headfirst into the world of MLOps, reinforcement learning, and neuro-symbolic systems.
Today, as a GenAI Solutions Lead at ABN-AMRO's CISO, I'm at the forefront of applying advanced AI to solve critical cybersecurity challenges. I architect and deploy secure, scalable Generative AI systems to detect sophisticated threats, from advanced phishing campaigns to automated secret scanning in vast codebases. My work is about building a digital immune system—intelligent, adaptive, and resilient.
My core mission is to translate state-of-the-art research into tangible, high-impact solutions. I believe the most powerful AI is built at the intersection of technical excellence and a deep understanding of the problems we aim to solve.
Experience
Senior Machine Learning Engineer · ABN-AMRO
As GenAI Solutions Lead within the CISO, I develop groundbreaking solutions for critical security challenges, including threat intelligence modeling, log normalization, and automated analysis of SOC II reports. I've enhanced multiple AI projects for threat detection (beaconing, phishing, secret scanning) and significantly improved the MLOps architecture for our security-focused ML systems.
- GenAI
- MLOps
- Cybersecurity
- PyTorch
Machine Learning Engineer · Amgen
Created an end-to-end multi-model MLOps pipeline using Databricks, MLFlow, and AWS for a computer vision system with 40+ cameras, enabling automated line clearance anomaly detection in under 30 seconds. Led ML model development for supply chain optimization and developed HMI for robotic systems. My contributions were recognized with 1st and 3rd place Innovation Awards.
- MLOps
- Computer Vision
- Databricks
- AWS
Research Intern · Philips Research
Developed first-of-a-kind Neuro-Symbolic (NeSy) architectures for Medical AI, utilizing symbolic constraints with temporal logic for my master's thesis on sepsis and emergency care treatment.
- Neuro-Symbolic AI
- Reinforcement Learning
Research Assistant & Co-ordinator · Vrije Universiteit
Spearheaded the creation of state-of-the-art interactive reinforcement learning algorithms that allow human experts to instruct agents in acquiring optimal policies in dynamic and complex scenarios.
- Interactive RL
- Human-in-the-loop
Research Intern · TNO
Pioneered an adaptive instructional support system using hybrid AI techniques (Neu-rules, Ontologies, Knowledge Graphs) to significantly improve employee learning outcomes.
- Knowledge Graphs
- Ontologies
Machine Learning Intern · Viroteq B.V.
Innovated a distinctive stability algorithm in Deep Reinforcement Learning to solve the 3D Bin Packing Problem for robotic palletization in warehouse environments.
- Deep RL
- Robotics
Research Intern · Vrije Universiteit
Contributed to Computational-AI Modeling by researching the development of human mental models as part of a master's thesis in Organizational Psychology.
- Cognitive Modeling
Education
MSc, Artificial Intelligence (Cum Laude)
Vrije Universiteit Amsterdam
MSc, Work & Organizational Psychology
Maastricht University
MA, Clinical Psychology (Distinction)
Fergusson College
BA, Psychology (Distinction)
Modern College
Projects
Wise-Vision
A production-grade MLOps pipeline for real-time anomaly detection on pharmaceutical assembly lines. This system uses computer vision and deep learning to monitor over 40 cameras, ensuring line clearance in under 30 seconds.
TeachME
An interactive reinforcement learning framework that allows human experts to teach AI agents optimal policies in complex, dynamic environments.
Robo Swimmers
A research project on evolutionary robotics, exploring the effects of aquatic and terrestrial conditions on the evolved morphologies and behaviors of simulated robots.
Skills
Research
A selection of my peer-reviewed publications. My work, comprising over 20 papers and 80+ citations, focuses on cognitive science, neuro-symbolic AI, and reinforcement learning.
Dynamics, Adaptation, and Control for Mental Models Analysed from a Self-modeling Network Viewpoint
Bhalwankar, R., & Treur, J. (2022). In: Advances in Cognitive Systems Engineering. Springer.
'What if I Would Have Done Otherwise…': A Controlled Adaptive Network Model for Mental Models in Counterfactual Thinking
Bhalwankar, R., & Treur, J. (2022). In: Brain-Inspired Cognitive Architectures for Artificial Intelligence. Springer.
An Adaptive Self-Modeling Network Model for Scenario-Based Shared Mental Model Configuration within Organisational Learning
Bhalwankar, R., & Treur, J. (2021). In: Procedia Computer Science, vol. 192. Elsevier.
Modeling the Development of Internal Mental Models by an Adaptive Network Model
Bhalwankar, R., & Treur, J. (2020). In: 11th Int. Conf. on Brain-Inspired Cognitive Architectures for AI (BICA*AI'20).
Get In Touch
I'm always open to discussing new projects, research collaborations, or interesting opportunities in the world of AI. Feel free to connect with me.
Connect on LinkedIn