AI Engineer & Technical Operator

Hi, I'm Adrian
Building AI Systems

Specializing in multi-agent workflows, local automation, and production-grade AI solutions

I build and ship AI-powered software: local agent workflows, healthcare transaction search, franchise operations tools, and business automation systems that survive real production use.

Proof I ship

Production healthcare search: full-stack Availity Transaction Search for finding individual healthcare transactions and their associated ecosystem data.
Production operations platform: led RoomComm from planning and architecture through shipped code for franchise room-status communication.
Local AI agent OS: built Hermes workflows with Qwen workers, Kanban task graphs, browser CDP automation, and human approval gates.
Measured automation: business workflows that reduce repetitive operational work by 60%+ while keeping humans in control.

View AI Systems Get in Touch

agent_orchestrator.py

class HermesAgent:
  def __init__(self):
    self.workflow = "Kanban-driven"
    self.safety = HumanInLoop
    self.orchestration = "Multi-Agent"
# Agent nodes with specialized roles
  nodes = [
    "Project Manager", "Triage",
    "Builder", "Ops"
  ]

● PM

● Triage

● Builder

● Ops

About

Software engineer & technical operator

I'm a software engineer and technical operator building practical AI systems for real-world workflows. With 7+ years of enterprise software experience and active work in local agents, LLM applications, and automation, I bridge the gap between cutting-edge AI research and production-grade solutions.

My focus: designing and deploying multi-agent systems (like Hermes), automating business processes with AI, and building tools that help teams work smarter—not just faster. I believe in human-in-the-loop safety, local-first AI where possible, and systems that actually ship.

Skills & Expertise

Technical capabilities across the AI stack

AI & LLM Systems

LLM Applications Multi-Agent Systems Prompt Engineering Workflow Design LLM Evaluation RAG Systems Function Calling

Automation & Backend

Python API Design Data Workflows Browser Automation (CDP) Task Orchestration Event-Driven Systems SQLite / PostgreSQL

Cloud & DevOps

AWS Docker GitHub Actions CI/CD Pipelines Local-First Architecture System Design Monitoring & Observability

Product & Business

Business Process Automation Product Thinking UX Design Technical Writing Stakeholder Management Requirements Analysis

Selected AI Systems

Case studies in applied AI and automation

🤖

AI Agent Systems

Hermes Local Agent OS

Multi-agent workflow system with human-in-the-loop safety gates. Features Kanban integration, browser automation via CDP, and local Qwen workers for autonomous task execution.

Python LLM (Qwen) CDP SQLite

🏥

Health AI & Analytics

Oura Health AI Analysis

Personal health data pipeline with AI-assisted trend detection for sleep, recovery, and activity. Privacy-first local processing with actionable wellness insights.

Python Oura API Pandas Matplotlib

💰

Finance Automation

Budget & Finance Automation

AI-powered transaction categorization and budget analysis with real-time tracking, predictive forecasting, and anomaly detection.

Python LLM Classification Plaid API Plotly

⚡

Business Automation

Business Automation Systems

Automated workflows for multiple clients including email drafting, customer service, and operational tasks. Reduced manual workload by 60%+.

Python LLM APIs Automation

🔎

Healthcare Transaction Search

Availity Transaction Search

Google Search for healthcare transactions: a full-stack app for finding individual transactions and their associated data across the healthcare ecosystem.

Full Stack Healthcare Data Production

💬

Communication Platform

RoomComm

Complex room-status communication platform for franchise locations, improving coordination, faster checkouts, and better customer service.

Full Stack Architecture Production

Adrian Santos