Hello, I'm

Amirreza Alasti

AI Engineer · Agentic Systems · LLM Agents

Incoming Ph.D. researcher at the L3S Research Center. Scroll to look closer: into the eye, through a neural network, and into the space of what I know and what I enjoy working on.

It starts with how we see

AI Engineer focused on agentic LLM systems, from context engineering and knowledge-graph memory to guardrails and deployment. M.Sc. in Computer Engineering (grade 1.0), first author at CIKM 2026 and DAI 2025, and an incoming Ph.D. researcher at the L3S Research Center.

Trained on experience

Inputs: the places that shaped me. Hidden layers: years of learning. Outputs: the fields I work in today. Every forward pass below runs left to right.

1. 2019B.Sc. Computer Engineering, Ferdowsi University of Mashhad
2. 2022Software Engineer & Data Science Researcher, Samian Soft
3. 2024M.Sc. Computer Engineering, Leibniz University Hannover (grade 1.0)
4. 2024Student research assistant, TIB Data Science & Digital Libraries
5. 2025AI speech intern, Sonova · Deutschlandstipendium · Dean's List
6. 2026Master's thesis on LLM web agents, Volkswagen AG · 1st place BBDC
7. Dec 2026Ph.D. Agentic AI Systems, L3S Research Center

Agentic Systems

LLM agents that perceive, plan and act, with context engineering, memory and humans in the loop.

• LLM web agent for multi-step workflows

  Volkswagen AG · Apr 2026 - Oct 2026

  Designed and built an LLM web agent for multi-step workflows using agentic context engineering and selective multimodal reasoning, reaching a strict success rate of 69.7% to 95.9% across multiple benchmarks.

  Thesis
• WCI: The Web Context Interface

  ACM International Conference on Information and Knowledge Management (CIKM 2026) · Accepted

  A. Alasti, N. Ghandeharioun, O. Karras

• TIB AIssistant

  TIB · IDaS

  LLM-based assistant that supports researchers across the entire research and writing process with source-grounded answers.

  AIssistant arXiv

Knowledge Graphs

Structuring scientific knowledge so that people and machines can query, reuse and reason over it.

• EmpiRE-Compass

  Technical lead of 5 students · REFSQ 2026 Best Posters & Tools Award

  Neuro-symbolic dashboard for sustainable and dynamic exploration, synthesis and reuse of research knowledge.

  Dashboard arXiv
• OntoAligner

  ESWC 2025 Best Resource Paper Award

  Modular Python toolkit for ontology alignment; contributed to development and documentation.

  GitHub
• NLP4RE ID cards → ORKG

  Joint Proceedings of REFSQ-2026 Workshops (NLP4RE) · CEUR-WS Vol-4208

  Pipeline that turns PDF ID cards into structured, FAIR knowledge in the Open Research Knowledge Graph.

  Paper
• ORKG Atlas

  Demo · AIKG-SD 2026 summer school

  Neuro-symbolic dashboard for exploring ORKG templates.

  Demo

Machine & Reinforcement Learning

Making models learn, and learn in the right order: from curriculum RL to winning data challenges.

• Learning to Play Blackjack: A Curriculum Learning Perspective

  International Conference on Distributed Artificial Intelligence (DAI 2025), London · Oral presentation

  LLM-guided curriculum learning for multi-agent reinforcement learning.

  arXiv Code
• Bremen Big Data Challenge 2026: 1st place

  Detection of age-related diseases

  Machine learning pipeline for disease detection from multimodal data.

  Solution

Signals & Speech

Neural audio: reconstructing speech that in-ear devices barely capture.

• Intern – AI-Based Speech Signal Processing

  Sonova · Oct 2025 - Mar 2026 · Hannover, Germany

  Bandwidth extension of occluded in-ear speech signals (earbuds, headbands) via neural reconstruction.

• Curated and recorded a custom speech dataset; developed data augmentation strategies for different noise sources.
• Implemented, trained and compared real-time-capable models with PyTorch, Librosa and SciPy.

Software Engineering

Shipping it: full-stack products, pipelines and the infrastructure under the models.

• Software Engineer & Data Science Researcher

  Samian Soft · Apr 2022 - Jul 2024 · Mashhad, Iran

  Built AI-driven automation for process management in ERP and CRM systems (React, Next.js, .NET).

• Lead developer of point-of-sale (POS) software as progressive web apps (PWA).
• Bachelor's thesis (grade 1.0): optimized visitor and freight distribution by combining routing algorithms (VROOM) with clustering methods for shorter routes and higher efficiency.

Teaching & Community

Sharing what I learn: teaching, mentoring, open science and the AI community.

• Teacher · i3center

  Dec 2023 - Aug 2024

  Mentored over 100 students in Machine Learning, Deep Learning, Data Analysis, and Software Engineering. Led interactive workshops, trained students in end-to-end project development, and organized comprehensive exams to assess their progress.

• Teaching Assistant, Ferdowsi University of Mashhad

  Six B.Sc. courses, including Artificial Intelligence, Databases and Software Engineering.

• Area Chair (Meta-Reviewer) · International Conference on Distributed Artificial Intelligence (DAI 2026)

  2026

  Website
• Participant · Summer School AI and Knowledge Graphs for Scholarly Data (AIKG-SD 2026)

  2026

  Website
• Member · AI Grid

  2025
• Stage Staff · WeAreDevelopers World Congress

  2025

The Frontier

Still training: agentic AI systems and representation learning with JEPA, starting December 2026.

• Ph.D. – Agentic AI Systems

  L3S Research Center, Leibniz University Hannover · Starting Dec 2026

  Research on agentic AI systems and their foundational methods, from training agents to representation learning with JEPA (Joint-Embedding Predictive Architectures). SciJEPA: JEPA embeddings for scientific knowledge graphs; supervised by Prof. Dr. Sören Auer.

  L3S

Converged, for now

The signal has reached every region, but the frontier keeps moving. If you'd like to explore it together, reach out.