🔭 I'm currently working on AI-powered workflow systems and operational automation using Python, OpenAI API, and Streamlit.
🌱 I'm currently learning AI engineering and neural networks through IBM's certification program.
👯 I'm looking to collaborate on AI workflow automation, ops tooling, and intelligent system design.
💬 Ask me about building AI applications with state management, LLM orchestration, and structured data extraction.
📫 How to reach me: mg.systems.dev@gmail.com

Service and operations manager with 10+ years leading field teams and 2+ years designing and shipping AI-powered workflow applications using Python, the OpenAI API, and prompt engineering. I build fully operational AI systems with LLM orchestration, state management, workflow automation, and structured data extraction.

Currently managing make-ready operations across 1,244 units at Billingsly while building AI workflow systems and data automation tools. Previously at MAA (535+ units), where I designed Python-based automation that saved $25K+ annually, reduced manual oversight by 40%, and trained 50+ staff in adopting new workflows.

Event-driven personal assistant with two-phase LLM orchestration → Phase 1: conversational reply | Phase 2: silent extraction of tasks, goals, time logs, and notes → Event-based state tracking converting conversations into queryable system state → Orchestration handling conversation memory, rolling summarization, and JSON parsing → Python, OpenAI API, SQLite, Streamlit

State-driven learning system with guarded progression → Structured programming curriculum with concept-locking logic → Dynamic prompt pipelines assembling instruction based on learner state and progress → Session persistence with defensive parsing and validation fallback → Python, OpenAI API, SQLite, Streamlit/Flet

AI practice environment with structured evaluation → Generates exercises and evaluates responses using structured JSON outputs → Defensive parsing pipelines with validation fallback and parser logging → Python, OpenAI API, SQLite, Streamlit/Flet

Relational knowledge management and operational planning → Complex relational databases with formulas and automation pipelines → Project tracking, operational planning, and personal productivity

End-to-end operational analytics pipeline → Python + Pandas ETL pipeline transforming turnover data into lifecycle metrics, task schedules, and SLA alerts → Streamlit dashboard reducing reporting turnaround by ~50% across 535+ units → Used by 20+ operational users for daily coordination

Guided learning lab for SQL and data transformation → Relational schema with seed data and query patterns → Joins, aggregations, filtering, and dataset merging

📁 Source Code: data-portfolio | Live App: Streamlit

• 🎓 BBA in Computer Information Systems — Ana G. Méndez University (In Progress)
• 🧾 Certifications:
  • IBM AI Engineering Certificate (In Progress)
  • Python for Everybody — University of Michigan, 2025
  • Google Project Management Certificate — Coursera, 2025
  • EPA 608 Certification — HVAC, 2018