Software and AI/ML engineer from Loja, Ecuador. I build LLM evaluation pipelines, RAG agents, data platforms and full-stack apps. Then I measure them properly, because looking like it works isn't the same as working.
- AI Evaluation Specialist for LinkedIn, contract via GreenLight (since Apr 2026). Benchmark datasets and annotations for code-generation and code-editing models.
- Final year of Computer Science Engineering at UTPL. My thesis trains a spiking neural network built from the fruit-fly connectome. Code in
flywire-snn-stdp.
- Scale AI, AI Research & Evaluation Engineer (Jul 2025 – Jul 2026). LLM evaluation pipelines, RLHF/RLAIF, MCP compliance checks.
- Clipp, Machine Learning Engineer, part-time (Oct 2024 – Aug 2025). Analytics for six product lines and user segmentation.
- Clínica Hospital San José, Full Stack Developer (Oct 2024 – Feb 2025). Clinic management system in React and Node, deployed and used in production.
- Insurance Policy AI Assistant. Multi-tool agent with a FAISS RAG layer, Gemini and web search. Answers cite their sources and retrieval is scored with RAGAS. I built the agent core and the retrieval pipeline.
- Clipp Analytics Platform. KPI dashboards for six service lines. I designed the architecture and the ETL, with Airflow loading PostgreSQL and FastAPI on top.
- SIGMA-IA Clipp. User segmentation for the Clipp Events app with K-Means. I led it.
- flywire-snn-stdp. Connectome-derived SNN with STDP, tested against a degree-matched random control.
- ML / AI · Python, scikit-learn, PyTorch, TensorFlow, LangChain, FAISS, RAGAS, Brian2
- Software · JavaScript, React, Next.js, Node.js, FastAPI, SQL
- Infra · Docker, Airflow, PostgreSQL, AWS, Git
Open to AI/ML and software engineering roles in Spain, the US or remote. Write to ljmora004@outlook.com