Legal Engineer | Applied AI & NLP Developer | Python Specialist
I bridge the gap between strict legal compliance and advanced software engineering. Currently based in Hokkaido, Japan, I specialize in building deterministic AI systems, automating complex legal workflows, and developing scalable Trust & Safety solutions. My goal is to translate rigid legal dogma into flawless, executable code.
- Core Engineering: Python (Pandas, Regex), API Integration, Deterministic Logic Routing.
- Applied AI: Large Language Models (LLMs), Prompt Engineering, Zero-Entropy Data Extraction, NLP.
- Systems & DevOps: Linux, Bash, Git, Server Deployment.
- Domain Expertise: LegalTech Architecture, Contract Law, AML (Anti-Money Laundering), Fraud Detection, C2C Marketplace Compliance.
- What it is: A neuro-symbolic AI framework designed to eliminate LLM hallucinations in enterprise contract review.
- How it works: It aggressively decouples generative NLP data extraction from decision-making. It uses zero-temperature LLMs solely to parse unstructured legal documents into JSON, and routes that data through a hard-coded, deterministic Python logic graph to execute flawless, 100% auditable compliance decisions.
🛡️ [SlangHunter]
- What it is: A semantic risk detection engine (MVP) built for C2C marketplaces.
- How it works: Deployed to automate Trust & Safety operations, it utilizes advanced NLP and Regex to instantly detect drug slang, money laundering patterns, and 'surikae' (bait-and-switch) fraud attempts, drastically reducing manual review times.
- Small Language Models (SLMs) for local edge deployment.
- Ensemble Machine Learning (LightGBM/XGBoost) for predictive legal analytics.
- LinkedIn: in/luis-legal-engineer
- Email: luisdossantos2001@gmail.com