Stanford University
Stanford Research

Optimas: Optimizing Compound AI Systems

Research collaboration at Stanford University on optimizing compound AI systems with globally aligned local rewards, advancing multi-component AI system performance.

Compound AI Multi-Agent Systems Optimization Stanford
Zopyros AI
Business

Zopyros Solutions

Founded AI transformation consultancy for medium to large enterprises. Helping companies harness the full potential of AI through strategic implementation, custom solutions, and cutting-edge technology integration.

AI Transformation Enterprise Solutions Strategic Implementation Entrepreneurship
ASEF Fellowship
Fellowship

ASEF Junior Fellow

Selected as Junior Fellow at the American Slovenian Education Foundation. Fellowship program supporting academic excellence and fostering collaboration between Slovenian and American academic communities.

Fellowship International Collaboration Academic Excellence Research Support
Big Bang AI Project
Business

Big Bang AI Product Transformation

Completed project helping Big Bang leverage the power of AI with their product catalog. Implemented AI-driven solutions to transform their product management and customer experience.

AI Implementation Product Catalog Enterprise AI Completed Project
Dormeo AI Project
Business

Dormeo AI Solutions

Completed AI implementation projects for Dormeo helping them in their fast growth. Delivered custom AI solutions to enhance their customer experience through advanced technology integration.

AI Solutions Custom Implementation Customer Experience Completed Project
Adrian with Stephen Wolfram
Research Program

Wolfram Summer School

Selected to work directly with Stephen Wolfram on Multi-Agent Simulations with Large Language Models. Developed complex simulations to study emergent behaviors in LLM-agent societies.

LLM Agents Complex Systems Wolfram Language Multi-Agent Simulation
Adrian with IHHN Carnegie Mellon University Project Group
Healthcare AI

Carnegie Mellon IHHN Healthcare AI

AI Intern for Healthcare Applications at Carnegie Mellon University. Worked on AI for Social Good initiative with deployment at IHHN Hospital, Pakistan, mentored by Prof. Rayid Ghani.

Healthcare AI AI for Social Good LLM Applications International Deployment
Predicting Impactful Events with Graph ML
Research

Predicting Impactful Events with Graph ML

Research project focused on predicting impactful events using Graph Machine Learning techniques. Advanced methodologies for event prediction and impact analysis using graph neural networks.

Graph ML Event Prediction Graph Neural Networks Impact Analysis
Political Bias
Research

Political Bias Detection in News

Language-agnostic framework for detecting political bias in multilingual news headlines. Published in Knowledge-Based Systems.

NLP Bias Detection Multilingual BERT
SLOmet
Research

SLOmet - Slovenian Commonsense Description

Slovenian adaptation of commonsense reasoning models, extending multilingual commonsense capabilities to Slovenian language with improved cultural and linguistic context understanding.

PyTorch Transformers Slovenian NLP Commonsense Reasoning
Adrian speaking at European Youth Parliament
Youth Leadership

European Youth Parliament Delegate

Participated in Thessaloniki International Forum'21, the 7th International Forum of EYP Greece. Engaged with 170+ young Europeans in discussions about the Future of Europe and societal challenges.

Youth Leadership European Policy International Dialogue Cultural Exchange
AT-IRCAI
Social Impact

Assistive Technology Tracker

Real-time tracking system for assistive technologies, developed with UCL and UNESCO IRCAI for global accessibility monitoring.

Angular D3.js MongoDB Node.js
Agent Models
Research

Agent-Based Text Understanding

Novel approach to NLP using agent-based models for semantic comprehension with improved interpretability.

Multi-Agent Systems PyTorch NetworkX NLTK
MultiCOMET
Full Stack AI

MultiCOMET - Multilingual Commonsense

Interactive web application for multilingual commonsense descriptions. Deployed after publishing the MultiCOMET paper, supporting 50+ languages.

PyTorch Transformers Angular NLP
Predictive Science
Research

Predicting Scientific Breakthroughs

Machine learning system analyzing research trends to predict emerging breakthroughs 2-3 years in advance.

TensorFlow Graph Neural Networks Time Series Big Data

Publications

2025

Optimas: Optimizing Compound AI Systems with Globally Aligned Local Rewards

Shirley Wu, Parth Sarthi, Shiyu Zhao, Aaron Lee, Herumb Shandilya, Adrian Mladenić Grobelnik, Nurendra Choudhary, Eddie Huang, Karthik Subbian, Linjun Zhang, Diyi Yang, James Zou, Jure Leskovec

arXiv Preprint

2025

Towards AI-Powered Real-Time Negotiation Agent

Martin Žust, Marko Grobelnik, Adrian M. Grobelnik, Abdul Sittar

SemGenAge Workshop at ESWC 2025

2025

Automation of Slovenian Honey Pollen Analysis

Adrian Mladenić Grobelnik, et al.

Specialized AI Research

2023

A commonsense-infused language-agnostic learning framework for enhancing prediction of political bias in multilingual news headlines

S Swati, Adrian Mladenić Grobelnik, Dunja Mladenić

Knowledge-Based Systems, Elsevier

2023

Emergent Behaviors from LLM-Agent Simulations

Adrian Mladenić Grobelnik

SiKDD 2023 Conference

2022

SLOmet – Slovenian Commonsense Description

Adrian Mladenić Grobelnik, Erik Novak, Marko Grobelnik, Dunja Mladenić

SiKDD 2022 Conference

2021

Understanding Text Using Agent Based Models

Adrian Mladenić Grobelnik

SiKDD 2021 Conference

2020

MultiCOMET - Multilingual Commonsense Description

Adrian Mladenić Grobelnik, et al.

EMNLP Findings

2019

Predicting The Next Big Thing In Science

Adrian Mladenić Grobelnik

SiKDD 2019 Conference

Let's Connect

Available for research collaborations, AI consulting, and innovative projects