Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I'm Uliuliuc Serafim, a Data Engineer and Machine Learning Engineer with a solid foundation in Artificial Intelligence. I build scalable data systems, ETL pipelines, AI/ML solutions, and backend services using Python, SQL, and modern ML frameworks. I love turning complex requirements into reliable, efficient technical solutions. From prototyping with synthetic data in hackathons to delivering production-grade components and guardrails for safe AI outputs, I enjoy collaborating across domains to drive actionable insights.…I'm Uliuliuc Serafim, a Data Engineer and Machine Learning Engineer with a solid foundation in Artificial Intelligence. I build scalable data systems, ETL pipelines, AI/ML solutions, and backend services using Python, SQL, and modern ML frameworks. I love turning complex requirements into reliable, efficient technical solutions. From prototyping with synthetic data in hackathons to delivering production-grade components and guardrails for safe AI outputs, I enjoy collaborating across domains to drive actionable insights.

Uliuliuc Serafim

AI Engineer, Data Scientist, Full Stack Developer, +2





I'm Uliuliuc Serafim, a Data Engineer and Machine Learning Engineer with a solid foundation in Artificial Intelligence. I build scalable data systems, ETL pipelines, AI/ML solutions, and backend services using Python, SQL, and modern ML frameworks. I love turning complex requirements into reliable, efficient technical solutions. From prototyping with synthetic data in hackathons to delivering production-grade components and guardrails for safe AI outputs, I enjoy collaborating across domains to drive actionable insights.…I'm Uliuliuc Serafim, a Data Engineer and Machine Learning Engineer with a solid foundation in Artificial Intelligence. I build scalable data systems, ETL pipelines, AI/ML solutions, and backend services using Python, SQL, and modern ML frameworks. I love turning complex requirements into reliable, efficient technical solutions. From prototyping with synthetic data in hackathons to delivering production-grade components and guardrails for safe AI outputs, I enjoy collaborating across domains to drive actionable insights.

Available to hire

I’m Uliuliuc Serafim, a Data Engineer and Machine Learning Engineer with a solid foundation in Artificial Intelligence. I build scalable data systems, ETL pipelines, AI/ML solutions, and backend services using Python, SQL, and modern ML frameworks.

I love turning complex requirements into reliable, efficient technical solutions. From prototyping with synthetic data in hackathons to delivering production-grade components and guardrails for safe AI outputs, I enjoy collaborating across domains to drive actionable insights.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Language

Romanian, Moldavian, Moldovan

Fluent

English

Fluent

Work Experience

Data Engineer / ML Engineer at Endava Iasi, Romania

May 1, 2025 - Present

Endava Healthtech Hackathon: Designed a proof-of-concept during a 2-day hackathon focusing on comparing standard clinical protocols with real-world patient care. Built a patient-journey model using synthetically generated data, enabling identification of protocol deviations and care patterns, and explored AI-assisted summarization of clinician notes to extract meaningful insights about patient care. AI Development & Testing: Built an extensible AI chatbot platform evolving from basic conversational capabilities to advanced RAG pipelines and multi-agent systems. Designed datasets and fine-tuned a BERT classifier for fraud detection. Implemented optimized context-chunking, retrieval ranking, and PDF-to-text processing pipelines. Evaluated LLM testing frameworks and contributed to high-quality Python codebases. Duplicate Patient Detector: Developed an ML pipeline for entity matching using embeddings, TF-IDF, similarity metrics, and clustering (DBSCAN, HDBSCAN, K-Means). Implemented blocki