Taras Andrushko
I am Taras Andrushko, an AI Engineer and Language Technology Specialist based in Oslo, Norway, with a strong background in conversational AI, Retrieval-Augmented Generation, and speech processing.
About
I am Taras Andrushko, an AI Engineer and Language Technology Specialist based in Oslo, Norway, with a strong background in conversational AI, Retrieval-Augmented Generation, and speech processing. I enjoy building end-to-end AI pipelines and integrating models into full-stack applications, focusing on reliability, evaluation, and prompt engineering.
I have Master’s in Language Technology at the University of Oslo (Aug 2023 – Dec 2025), researching automated detection of personally identifiable information in speech data, and I love collaborating across research and production to deliver privacy-aware, user-centric AI solutions.
- 4 years' experience
- Joined May 2026
Skills· 14
Languages
- EnglishFluent
- NorwegianAdvanced
- GermanIntermediate
- RussianFluent
- UkrainianFluent
Industry experience
- Software & Internet
- Education
- Professional Services
- Media & Entertainment
- Other
- Non-Profit Organization
Work experience· 7 roles
Designing and building a full-stack web application for learning Norwegian for professional purposes (Yrkesnorsk). Developing an AI-powered chatbot (Snakkis) using LLMs and RAG for contextualized guidance and learning support using OpenAI API. Implementing document ingestion, embedding, and retrieval pipelines to ground responses in curriculum-aligned content. Integrating a web-based chat interface enabling interactive dialogue, guided exercises, and domain-specific feedback (ReactJS, TypeScript). Applying prompt engineering and iterative evaluation to improve response quality, reduce hallucinations, and ensure linguistic accuracy.
Co-founded the project, contributed to technical strategy, and led product development initiatives. Built the initial architecture and spurred user growth and funding.
Built responsive, multilingual platforms with JavaScript, Python, TypeScript, .NET and F#; delivered accessible UIs; backend services using Python (Flask / FastAPI-style APIs); integrated external services (Google APIs) and secure authentication; implemented automated testing with PyTest and CI/CD pipelines; improved UX and deployment.
Education
Certifications and awards
Hire Taras Andrushko today
Post a job and invite Taras, or contact them directly. No hidden commission.