Natural Language Processing
Lithuanian language technology, social text analysis, corpus tools, sentiment, emotion and hate speech detection.
AI • NLP • LLMs • Full-stack development
Building practical AI systems and real-world workflows.
I am an IT professional, researcher, and Full-Stack Developer with more than a decade of experience in software engineering, data science, Natural Language Processing, deep learning, and large language models.
About
Research-minded engineer with a builder’s mindset.
My work combines technical architecture, applied AI research, and full-stack product development. I focus on systems that convert raw data, documents, audio, cybersecurity events, and Lithuanian-language texts into useful tools, models, dashboards, and decision support workflows.
I am especially interested in low-resource language technologies, social text analysis, hate speech detection, sentiment and emotion recognition, synthetic datasets, chatbots, and retrieval-augmented generation.
0+ years in software
0AI / NLP focus areas
0selected publications
Expertise
From AI prototypes to production-ready web systems.
Search the skills below or switch between areas.
Large Language Models
LLM workflows, RAG, agentic generation, chatbot development, evaluation loops, and prompt-driven systems.
Machine Learning & Deep Learning
Transformers, clustering, classification, data augmentation, model evaluation, and applied AI experiments.
Full-stack Development
JavaScript frontends, Java backends, APIs, dashboards, admin systems, databases, deployment and maintenance.
Data Pipelines
PDF-to-Markdown, OCR, text extraction, synthetic QA generation, data quality workflows, and corpus preparation.
Cybersecurity Analytics
IDMEFv2 analysis, malicious campaign detection, analyst dashboards, event exploration, and AI-generated insights.
Semantic Networks
Topic Maps, semantic networks, knowledge-enhanced applications, SQL and NoSQL-backed data modelling.
Academic Research
Conference presentations, publications, low-resource NLP, word embeddings, and Lithuanian language benchmarks.
No matching skills found.
Experience
Selected professional timeline
Filter by focus area and expand entries for more detail.
Developing AI-based solutions for cybersecurity analysis, including IDMEFv2 message analysis with machine learning, clustering, LLMs, RAG, and interactive analyst dashboards.
Built an AI-based system for converting Lithuanian texts into easy-to-read language using an agentic writer–critic workflow, repeated refinement, and support for multiple LLM providers.
Contributed to the design of a data generation pipeline for a large-scale Lithuanian question–answer dataset intended for chatbots and AI applications.
Developed an application for uploading audio/video files and receiving automatic transcriptions and summaries using WhisperX, Semantika, and Gemma 3.
Created PDF-to-Markdown software for corpus preparation, supporting scanned documents, machine-readable PDFs, OCR processing, text extraction, and review workflows.
Led technical architecture and development for AI-driven digitization, process automation, service optimization, and specialized chatbot solutions.
Researched language technology applications, including sentiment analysis, emotion detection, hate speech identification, and transformer-based NLP models.
Full-stack software development, including JavaScript frontends, Java backends, web applications, software configuration, improvement, and maintenance.
Developed language technology tools for Lithuanian, including NLP and corpus tools, while contributing to specialized linguistic research systems.
Built Java, Spring, Vaadin, semantic network, frontend, backend, SQL and NoSQL-backed web applications.
Projects
Selected systems and applied AI work
Cybersecurity AI
SAFE4SOC AI analytics
IDMEFv2 analysis with ML, clustering, LLMs, RAG, and a dashboard for exploring attacks and AI-generated insights.
Language accessibility
Easy-to-read Lithuanian
Agentic system for converting standard Lithuanian texts into clearer easy-to-read language using writer and critic agents.
Data pipeline
Lithuanian QA corpus
Pipeline design for transforming public Lithuanian information into synthetic question–answer pairs for AI applications.
Document AI
PDF to Markdown workflows
Software for extracting, reviewing, correcting, and preparing text resources from scanned and digital PDF documents.
Research
Low-resource NLP, social text, and Lithuanian language models.
Current doctoral research focuses on automated creation and application of Lithuanian social-text language models in deep learning and NLP tasks.
- Exploring Hate Speech Detection Models for Lithuanian Language, WOAH 2025.
- Fine-tuning language models for emotion recognition in Lithuanian texts, CEUR Workshop Proceedings, 2023.
- Extrinsic evaluation of word embedding models using semisupervised NLP tasks, DAMSS 2022.
- Knowledge-base enriched word embeddings for social media text, DAMSS 2021.
- Intrinsic Word Embedding Model Evaluation for Lithuanian Language, IVUS 2021.
- DAMSS 2024 — poster presentation on Lithuanian hate speech detection models.
- University of Alicante, 2023 — emotion recognition in Lithuanian texts using transfer learning and NMT.
- DAMSS 2022 — word embedding model evaluation using semisupervised NLP tasks.
- DAMSS 2021 — knowledge-base enriched word embeddings for social media text.
- ACDL, Siena — Advanced Course on Data Science & Machine Learning.
Education
Education & training
Doctoral Student in Informatics
Vytautas Magnus University, Faculty of Informatics
Master’s Degree in Computer Science
Kaunas University of Technology, Faculty of Informatics
Bachelor’s Degree in Computer Science
Kaunas University of Technology, specialization in computer networks
Contact
Let’s build useful AI and language technology systems.
Available for research collaboration, AI/NLP systems, applied LLM workflows, dashboards, and full-stack development.