Type to search. Arrow keys to move, Enter to open, Esc to close.

Software Developer - AI & Backend

Tóth Dorina Ildikó

I work at the intersection of AI and software engineering. I am interested in not only using AI technologies but understanding them, and integrating them into reliable software systems that solve real problems.

Selected projectsHow I work
Tóth Dorina Ildikó

Focus areas

I work with LLMs, RAG systems, embeddings, prompt engineering, fine-tuning, document processing and Python-based AI applications, while also deepening my understanding of the foundations of machine learning and deep learning.

LLM and GenAI engineering

LLM-based systems, prompt engineering, fine-tuning, generative AI applications, and experiments with AI agents and agent frameworks.

RAG and embeddings

Retrieval-augmented systems, embeddings, chunking and vector search with PostgreSQL and pgvector.

Document intelligence

OCR with Tesseract, AI vision, document cleaning and content extraction, AI-based text processing.

NLP, ML and deep learning

Natural language processing, neural networks, computer vision, model training, and the mathematical foundations behind them.

Backend and APIs

Python and Django backends, API-based services, SQL and PostgreSQL, containerization with Docker.

Optimization and system design

Constraint satisfaction and SAT-based optimization, linear and logical constraints, software architecture.

About

My interest in AI began during my university years, when I started experimenting with the first LLMs and chatbots that could communicate in natural language and help with programming tasks. At first this steered me towards practical uses of NLP and generative AI; later I became more interested in what happens behind the technology. In Python I worked on machine learning and image processing projects, training models and working with neural networks and computer vision, and alongside them I studied how machine learning, deep learning, algorithms and the underlying mathematics work. In my thesis I studied the practical use of language models and prompting techniques.

I currently work as an AI developer. Besides building LLM-based systems, RAG, embeddings, prompt engineering, fine-tuning, document processing, OCR, AI-based text processing and Python systems, my work also covers classical software engineering and optimization problems. I am particularly interested in integrating AI capabilities into reliable, well-designed software systems, so I want to understand the technologies behind them, not just use them. In the coming years I want to grow towards an AI Engineer / Software Engineer role where I can independently design and build complex AI-based systems while keeping a research and experimental mindset.

For me, AI should not replace engineering thinking but serve as its tool. I use LLMs and vibe-coding-style tools mainly for faster prototyping, learning, experimenting and supporting the development process, while I consider it important to truly understand the code, the architecture and the technologies used. I like to examine a problem from several angles, experiment with different solutions, and then turn the working result into a real software system.

Soft skills

Experience

  1. 2025 - present

    Software Developer

    Eszterházy Károly Catholic University, IT Development Department

    Design, development and integration of Python-based systems that use AI, including model training and optimization. Technology areas: LLMs, RAG and embeddings, prompt engineering, fine-tuning, document processing and OCR (Tesseract, AI vision), AI-based text processing, and constraint-based optimization (SAT solving, linear and logical constraints). Experimenting with agent frameworks. Technologies: Python, PostgreSQL, Django.

  2. 2023 - 2024

    Teaching Assistant (Demonstrator)

    Eszterházy Károly Catholic University

    Supported three practical courses. High-Level Programming Languages I: procedural programming in C# (types, variables, operators, selection and iteration, arrays, lists, dictionaries and records, text file handling, pseudo-random numbers, methods). High-Level Programming Languages II: object-oriented programming in C# (classes, constructors, inheritance, abstract classes, interfaces, LINQ, delegates and lambda expressions, exception handling, generics, GUI programming) and planning and implementing independent projects. Application Development and Project Lab I: agile and classical project methods (Scrum, Kanban), Git-based version control and task tracking, requirement and functional specifications, system plans, UML diagrams, prototyping, testing, documentation and software quality assurance. Technologies: .NET, C#, Git, GitHub, UML, Trello, Markdown.

  3. 2022 - 2025

    Web Developer and Maintainer

    Kepes György College for Advanced Studies; individual client (2022 - 2024)

    Website development, maintenance and design with HTML, CSS and JavaScript, including content updates and technical operation. At the College also head of the external communication working group, financial officer and secretary.

Education

  1. 2022 - 2025

    BSc in Software Engineering

    Eszterházy Károly Catholic University

  2. 2018 - 2022

    Computer System Maintenance Technician

    Váci SZC Petőfi Sándor Technical School

Stack

Technologies and areas I work with, grouped by type.

Languages

  • Python
  • Java
  • C#
  • TypeScript
  • JavaScript
  • SQL
  • PHP

Frameworks and libraries

  • Django
  • Spring Boot
  • .NET
  • Laravel
  • React
  • Astro
  • Tailwind CSS
  • Bootstrap
  • Jinja2
  • Pillow
  • HTML5
  • CSS3

Databases

  • PostgreSQL
  • MySQL
  • SQLite
  • MongoDB

DevOps and tools

  • Docker
  • GitHub Actions
  • Git
  • GitHub
  • Bash
  • Linux
  • Cypress
  • PlantUML
  • Trello
  • Markdown
  • LaTeX
  • YAML
  • pytest
  • Apache HTTP Server

AI and ML

  • Hugging Face Transformers
  • Ollama
  • PyTorch
  • LangGraph
  • pgvector
  • Tesseract OCR
  • OpenAI API
  • Anthropic API

Concepts

  • LLM and GenAI
  • RAG
  • Embeddings
  • Prompt engineering
  • Fine-tuning
  • NLP
  • Deep learning
  • Computer vision
  • AI agents
  • Constraint solving (SAT)
  • REST APIs
  • System design

How I work

Every project goes through the same stages, from the first idea to the published result. A readiness check decides when an idea is ready to be built, and research stays tied to concrete project needs. The Method page explains the stages and how languages, technologies and projects relate.

  1. 01

    Idea

  2. 02

    Planning

  3. 03

    Research

  4. 04

    Architecture

  5. 05

    Implementation plan

  6. 06

    Readiness check

    Decision point
  7. 07

    Implementation

  8. 08

    Result

  9. 09

    Publication

Read about the method

Certificates

Completed courses and certifications.

AWS Foundations: Machine Learning Basics

AWS Training and Certification · Issued: 24 September 2026

AWS Foundations: Machine Learning Basics

An overview of machine learning concepts, terminology and processes: what ML is, how it solves business problems, when a model is appropriate, and the phases of an ML pipeline.

  • Machine learning
  • Three categories of ML algorithms
  • Deep learning
  • ML pipeline phases
Open PDF
Fundamentals of Machine Learning and Artificial Intelligence

AWS Training and Certification · Issued: 24 September 2026

Fundamentals of Machine Learning and Artificial Intelligence

Foundations of ML and AI and how AI, ML, deep learning and generative AI relate. Covers training data types, learning paradigms, neural networks, the foundation model lifecycle, and AWS services that use AI.

  • AI, ML, deep learning and generative AI
  • Supervised, unsupervised and reinforcement learning
  • Neural networks
  • Foundation models: LLMs, diffusion and multimodal models
  • AWS AI services (Amazon Bedrock, Comprehend, Polly)
Open PDF
Fundamentals of Generative AI

AWS Training and Certification · Issued: 4 October 2026

Fundamentals of Generative AI

An introduction to machine learning and generative AI, including NLP, and to AWS AI/ML offerings such as Amazon SageMaker and Amazon Bedrock. Covers data preparation, model training, deployment and monitoring, and choosing an appropriate AI/ML solution for a use case.

  • AI and ML basics
  • Generative AI and NLP
  • Amazon SageMaker and Amazon Bedrock
  • Data preparation, model training, deployment and monitoring
  • Selecting AI/ML solutions for specific use cases
Open PDF