Where AI gets used
I automate the parts of a team's work that quietly slow it down.
I spot what's slow in how a team already works and build agents, skills, and MCP servers to run it. Consulting through Common Ground AI.
~4
Reference scripts replaced per change
In the team's shared agent library. Ask for a column and it writes the column, the audit-logging entry, and the history-table trigger in one step — the package that used to mean copy-pasting from three or four old scripts.
Read the writeup →~10 hrs
Per-system WCAG audit, validated
Down from a 50–70 hour manual audit. Scored against WAVE, the accessibility tool built on the WebAIM Million: 3 measured hours closed 85% of the score gap on the first system, moving its average page score from 6.3 to 7+ out of 10.
Read the writeup →59
Recruiter profile views in 90 days
Since building the system that keeps my resume, LinkedIn, and portfolio aligned: 91 total profile views in the last 90 days, 59 of them recruiters, and a role that moved my pay up roughly 70–80%.
Read the writeup →About
I'm an AI Automation Engineer, currently on contract with Cisco building agentic AI workflows. Before that, I spent three years standing up the AI function at RDE Systems from scratch: agents, MCP servers, and the infrastructure other tools plug into. What I like most is finding the part of a job that's slow, then building something that fixes it for good.
Teaching is where I get the most energy. I've mentored and guided people at pretty much every stage: kids, high schoolers, college students, working professionals. Helping someone go from stuck to confident with something new is one of my favorite parts of any project.
That mix of building things and teaching people is what led me to start Common Ground AI, a consulting practice that helps small business owners untangle their workflows and run things a little more smoothly.
What I can help with
Microsoft 365 agents
Teams and Copilot agents for inbox triage, ticket tracking, and recurring checks.
MCP servers
Give an AI assistant real, scoped access to your database, codebase, or files, with an audit trail.
Prompt libraries
Version-controlled prompts for the requests your team makes over and over.
Claude skills
Packaged skills that fold a repeated multi-step task into a single instruction.
Dashboards
KPI dashboards built straight from an export, without a BI license or a warehouse project.
Data analysis
Turn a messy export into a clean dataset and a clear answer to the question behind it.
AI enablement & platform engineering
License rollouts, champion networks, and the MCP infrastructure underneath them.
Resume, LinkedIn & portfolio alignment
One fact-checked source of truth across your resume, LinkedIn, and portfolio, instead of three documents that quietly drift apart.
Experience
-
Cisco (Contract)
Contract engagement supporting Cisco's internal agentic AI workflow tooling.
-
RDE Systems
Led the AI initiative: two production agents, the company's first three MCP servers, and the AI angle on the UI/UX rebuild.
-
Common Ground AI Consulting
Independent AI consulting for small businesses. Shipped commongrnd.ai and workflows for a growing content creator.
-
The Coder School, Short Hills, NJ
Designed and taught project-based curricula in React, Python, and Java for students ages 7–21.
-
MetroStar
Built Tableau dashboards on Census data for city planners, and reusable React components from the Tableau JS API.
Certifications & education
AI Fluency: Framework & Foundations
Anthropic · Apr 2026
Teaching AI Fluency
Anthropic · Apr 2026
Introduction to Model Context Protocol
Anthropic · Jul 2026
B.S. Computer Science
New Jersey Institute of Technology · 2023
Projects
Turned the manual work of keeping a resume, LinkedIn, and portfolio consistent into one repeatable, fact-checked system.
As of Sept 2026: 91 profile views and 59 recruiter views in the past 90 days
Built a custom Claude Skill for Pine Script development, then used it to ship a harmonic pattern detector and a dynamic support/resistance indicator.
Shared with a 3,000+ member Discord trading community
An idea that stalled for years on a data problem: no public API, and the fight pages had URLs too irregular to crawl. Built a custom scraper to pull and normalize a clean dataset, then a Flask app on top of it: browse upcoming cards, sort fights by how close the stats favorite is, or compare two fighters head-to-head on career striking and grappling numbers.
Writing
-
09/18/2026
Standardizing my resume, LinkedIn, and portfolio into one agent system
Agentic AI Claude Code Career Tools -
06/23/2026
Learning to ask for help, then learning to be asked Editor's pick
Mentorship Leadership -
07/09/2026
How I ended up building the company's first MCP server
MCP Python SQL Server -
07/11/2026
Turning the UI rebuild into a design system
CSS UI/UX Design Systems -
06/22/2026
How I built a harmonic pattern detector in Pine Script
Pine Script Harmonic Patterns
Contact
Full-time roles and contract work get the fastest reply, but a one-off technical question is fine too. For contract work, tell me the outcome you need and I'll say whether it's a fit.
A sentence about the role or the problem is plenty. Add a link or a job description if you have one.
It lands in my personal inbox and I answer it myself.
Thirty minutes on what you're trying to do and where it's stuck. By the end we both know whether I'm the right person for it.
Send a message
This goes to my personal inbox. Prefer email? joan.escamilla1@gmail.com