Aizaz Sharif

I am Aizaz Sharif, a computer Science Major, AI Engineer, and Researcher in Oslo, Norway. As a Machine Learning Engineer at AI Dev Lab, I build end-to-end LLM architectures while previously researching autonomous driving systems at the VIAS department of Simula Research Laboratory and the University of Oslo.

You can download my resume from here: Aizaz Sharif Resume

What I’m Building Now

At AI Dev Lab, I’m implementing a production-ready LLM-powered platform (app.saleskik.ai) for processing daily meeting recordings. This involves:

Developing custom RAG solutions using LangChain, LangGraph, and Pinecone
Building robust backend infrastructure with Python, Firebase, and Google Cloud services
Creating intelligent conversation processing pipelines for context-aware meeting summarization
Managing the full ML/LLM development lifecycle from prototyping to production

Scientific Contributions

My doctoral research (defended on September 2024) focused on “Testing the Safety and Robustness of Autonomous Cars in a Multi-agent Environment,” where I made three novel scientific contributions:

Adversarial reinforcement learning techniques for testing and improving autonomous systems
Systematic benchmarking methodologies for multi-agent RL driving policies
Reward modeling frameworks for finding uncertain state space without adversarial training approach

I’ve created and open-sourced platforms for testing multi-agent autonomous driving systems, gaining extensive experience with simulated driving environments and reinforcement learning.

Beyond autonomous driving, I have also performed research in medical imaging, Android security, and software engineering which resulted in multiple publications, including two conference papers and a journal article. This diverse research background has strengthened my ability to apply machine learning techniques across various domains.

Project Ownership & Team Leadership

I’ve successfully led teams and managed complex technical projects throughout my career:

At Simula, I independently managed my entire research portfolio, from literature review to conceptualizing novel ideas, executing experiments, analyzing results, and writing papers - all while working autonomously to deliver high-quality scientific contributions
At NCCS, I led a team developing a ‘Mobile Phone Digital Forensics’ toolkit, overseeing the entire project lifecycle from concept to beta testing with authorities
At AI Dev LAb, I managed end-to-end ML/LLM development workflows and coordinated with cross-functional teams to deliver production-ready AI solutions. I also collaborated directly with clients to understand business requirements and translate them into technical solutions

Life Outside Code

Outside of my professional life, I enjoy photography and staying active through cycling, table tennis, and squash. I’m an enthusiastic Tekken player trying to compete at a higher level, and occasionally play other games like Rainbow Six Siege when time allows. I have a curiosity for learning about different cultures and collecting random facts. Living in Norway, I’m exploring the country’s beautiful landscapes and diverse regions, as well as engaging in travel photography during my adventures.

Contact

For collaboration, future opportunities, or any queries regarding my current or past work, you can reach out using my email: aizazsharif@gmail.com and LinkedIn.

news

Apr 23, 2025	Moved to a new website theme. Now completing unfinished blogs (both technical and memoirs).
Sep 27, 2024	Defended my PhD thesis
Jan 11, 2023	Blessed with a baby boy!

selected publications

ReMAV: Reward Modeling of Autonomous Vehicles for Finding Likely Failure Events

Aizaz Sharif and Dusica Marijan

In IEEE Open Journal of Intelligent Transportation Systems (OJ-ITS), 2024

DOI PDF Code
Adversarial Deep Reinforcement Learning for Improving the Robustness of Multi-agent Autonomous Driving Policies

Aizaz Sharif and Dusica Marijan

In 2022 29th Asia-Pacific Software Engineering Conference (APSEC), 2022

Abs DOI PDF Code Website

Autonomous cars are well known for being vulnerable to adversarial attacks that can compromise the safety of the car and pose danger to other road users. To effectively defend against adversaries, it is required to not only test autonomous cars for finding driving errors but to improve the robustness of the cars to these errors. To this end, in this paper, we propose a two-step methodology for autonomous cars that consists of (i) finding failure states in autonomous cars by training the adversarial driving agent, and (ii) improving the robustness of autonomous cars by retraining them with effective adversarial inputs. Our methodology supports testing autonomous cars in a multi-agent environment, where we train and compare adversarial car policy on two custom reward functions to test the driving control decision of autonomous cars. We run experiments in a vision-based high-fidelity urban driving simulated environment. Our results show that adversarial testing can be used for finding erroneous autonomous driving behavior, followed by adversarial training for improving the robustness of deep reinforcement learning-based autonomous driving policies. We demonstrate that the autonomous cars retrained using the effective adversarial inputs noticeably increase the performance of their driving policies in terms of reduced collision and offroad steering errors.
Evaluating the Robustness of Deep Reinforcement Learning for Autonomous Policies in a Multi-Agent Urban Driving Environment

Aizaz Sharif and Dusica Marijan

In 2022 IEEE 22nd International Conference on Software Quality, Reliability and Security (QRS), 2022

Abs DOI PDF Code

Background: Deep reinforcement learning is actively used for training autonomous car policies in a simulated driving environment. Due to the large availability of various reinforcement learning algorithms and the lack of their systematic comparison across different driving scenarios, we are unsure of which ones are more effective for training autonomous car software in single-agent as well as multi-agent driving environments. Aims: A benchmarking framework for the comparison of deep reinforcement learning in a vision-based autonomous driving will open up the possibilities for training better autonomous car driving policies. Method: To address these challenges, we provide an open and reusable benchmarking framework for systematic evaluation and comparative analysis of deep reinforcement learning algorithms for autonomous driving in a single- and multi-agent environment. Using the framework, we perform a comparative study of four discrete and two continuous action space deep reinforcement learning algorithms. We also propose a comprehensive multi-objective reward function designed for the evaluation of deep reinforcement learning-based autonomous driving agents. We run the experiments in a vision-only high-fidelity urban driving simulated environments. Results: The results indicate that only some of the deep reinforcement learning algorithms perform consistently better across single and multi-agent scenarios when trained in various multi-agent-only environment settings. For example, A3C- and TD3-based autonomous cars perform comparatively better in terms of more robust actions and minimal driving errors in both single and multi-agent scenarios. Conclusions: We conclude that different deep reinforcement learning algorithms exhibit different driving and testing performance in different scenarios, which underlines the need for their systematic comparative analysis. The benchmarking framework proposed in this paper facilitates such a comparison.