AI Engineer · NLP Researcher
Anderson Morillo
I build production RAG systems and publish NLP research on retrieval, persuasion, and hallucination filtering.
VerbaNexAI Lab (UTB) · applied AI for industry and research.
- SemEval 2023–2025
- IEEE C3 2025
- VerbaNexAI Lab · UTB
Featured work
SGM — Meteorological dashboards (San Andrés Island)
Problem
Operators needed clear, station-level weather boards for day-to-day monitoring.
Approach
TimescaleDB ingestion, REST APIs, and web dashboards for multi-station IoT data.
Result
Live boards used across stations such as Brooks Hill, RSU, and Interaseo.
Peer-reviewed NLP that ships
Shared-task and conference work on problems that show up in real systems—relatedness, persuasion, hallucinations, and biomedical retrieval.
- SemEval systems for framing, relatedness, and hallucination spans (2023–2025)
- Hybrid PubMed retrieval + LLM pipeline for biomedical QA (IEEE C3 2025)
- Research rooted in VerbaNexAI Lab at Universidad Tecnológica de Bolívar
Notes you can reuse
Practical write-ups on tooling and methods—not just theory. Many posts are also in Spanish.
- RAG over technical documentation
- Model Context Protocol (MCP) for agent tooling
- Weirdness-based keyword extraction and unigram corpora
What I help teams ship
Applied AI and data systems—from document-grounded assistants to monitoring platforms—backed by research and production work.
- Conversational RAG and private/on-prem LLM setups
- Dataset research, evaluation, and creation
- AI architecture for environmental and IoT monitoring