Curriculum Vitae

My professional background and qualifications

Shosei Abe

Leányka utca 2, Eger, Hungary 3300

Mobile: (+36)70-771-6906 E-mail: syousei1999@gmail.com

PERSONAL SUMMARY

Computer Science graduate specializing in AI/ML with hands-on experience deploying LLM-based systems. Winner of ElevenLabs Global Hackathon (2026). Proficient in PyTorch, LangChain, and production ML pipelines. Seeking AI Engineer role to apply research-driven solutions to real-world problems.

SKILL SET

  • Core AI/ML: PyTorch, TensorFlow, HuggingFace Transformers, LangChain, RAG, Fine-tuning, RLHF
  • Engineering: Python, Docker, Kubernetes, AWS (SageMaker, Lambda), CI/CD, Git
  • Data: PostgreSQL, MongoDB, Airflow, Spark, ETL, Data Visualization (Plotly, Streamlit)
  • Web & Mobile: JavaScript, TypeScript, React, React Native, Node.js, Express
  • Additional: C#, .NET, Java, Rust, Kotlin, C++, PHP, Formal Methods (Z3, Coq/Lean)

WORK EXPERIENCE

Refonte Learning

Dec 2024-Jun 2025

Data Engineering Intern

  • Developed ETL pipelines and optimized database performance for large-scale data processing.
  • Worked with cloud-based data warehouses and implemented data visualization dashboards.
  • Deployed ML models into real-time analytics pipelines.

DENSO Corporation

Apr 2018-Aug 2022

Production Control

  • Managed supply chain operations for automotive manufacturing

EDUCATION

Eszterhazy Karoly Catholic University - Hungary

Expected 06/2026

Bachelor of Science: Computer Science, Specialization in Data Science & Artificial Intelligence

GPA: 4.60 / 5.00

Stipendium Hungaricum Scholarship, 2024 from Tempus public foundation

PROJECTS

AGI/AI Safety Research

Jul 2025-Present

  • Bachelor Thesis: Designing a Self-Improving Research Agent: A Practical Step Towards AGI with Formal Verification.
  • Built an LLM-based autonomous research agent with iterative self-improvement.
  • Integrated formal verification (Z3, Coq/Lean) for provable safety and correctness.
  • Implemented LangGraph prototype for recursive research, document/code generation, and feedback-driven optimization.

UrbanSpace Optimizer

Dec 2024-Current

  • Hungarian Startup University Program's Project.
  • Built an AI-powered real-time curbside optimization system using IoT sensors & AI analytics to reduce urban congestion, enhance traffic safety, and optimize mobility.

Friction - Global Hackathon Winner

  • Winner (1st Place) at ElevenLabs World Wide Hackathon 2026 (The world's largest AI hackathon).
  • Deployed conversational AI agent using ElevenLabs Voice API + GPT-4
  • Achieved 73% reduction in impulse purchases in user testing (n=150)
  • Tech stack: React Native, Node.js, LangChain, Pinecone vector DB
  • Processed 10K+ user conversations with <500ms latency

PUBLICATIONS

Abe, S. (2025). "Explainable Artificial Intelligence in Autonomous Systems: Navigating Transparency and Safety in Critical Decision Environments." [Under Review/Preprint available on request]

LANGUAGE

  • English: C1 – Advanced, CEFR (verified by EF SET Certificate)
  • Japanese: Native
  • German: B1 – Intermediate, CEFR
  • Hungarian: A2 – Elementary, CEFR

ADDITIONAL INFORMATION

Volunteer:

  • English Camp Assistant, Alive English School Summer Camp - 2017
  • Information Desk, International Startup Event - 2023

Awards:

  • 1st Place Winner, Elevenlabs World Wide Hackathon, Budapest (World biggest hackathon) - 2026
  • Winner, Builders Weekend Hackathon, Tokyo (with Kraken Technologies) - 2023
  • Finalist, OVAL (East Asia's largest business contest) - 2023
  • Finalist, GEIL (Japan's largest policy-making contest) - 2023
  • 6th Place, Junction X Budapest (Hungary's largest hackathon) - 2023

Scholarships:

  • Stipendium Hungaricum Scholarship (Fully Funded Scholarship) - 2024
  • Hungarian Startup University Program Scholarship (Top 200/3000) - 2025