ED PELAEZ
AI Automation Specialist
I turn repetitive tasks into autonomous systems — so you can focus on what matters.
12+
Automations Built
50+
Hours Saved Weekly
7+
Clients Served
PROJECTS
Automations I built that save time, cut costs, and scale operations.
A multi-tenant web portal for a medical billing company that prepares insurance claims for its clients. Each client uploads a monthly billing sheet in whatever shape their system exports, the portal checks every row against that client's own reference data, fixes what it can prove and flags what it can't, and hands the account manager a file ready to submit. Two sides, one source of truth: clients answer only the questions that are theirs, account managers see everything. Owned the whole front end and the checking engine behind it.
A deterministic n8n workflow that watches a driver-recruiting agency's Google Sheets hiring trackers and flags any file that's gone quiet, so account managers never have to manually scan a sheet to find what needs attention. No AI anywhere in the loop — a spreadsheet script stamps the moment a file is actually worked, and a twice-daily scheduled check does the rest. Built and validated end-to-end with a live test.
Seven connected n8n workflows that run an ops coordinator role for a global climate advisory firm — requests flow in through a form, deal stage changes auto-spawn task playbooks in Notion, a nightly scan flags overdue and stale tasks with direct comments, a meeting-booked trigger fires a 12-step research agent that delivers a formatted brief before every sales call, and a daily lost-deal agent scans the CRM, judges re-engagement with AI, and emails the owner Approve/Skip. Built and presented live to the Co-Founder & COO in under five hours.
Medical Billing — Claims Pre-Check Portal
Book for a live demoA multi-tenant web portal for a medical billing company that prepares insurance claims for its clients. Each client uploads a monthly billing sheet in whatever shape their system exports, the portal checks every row against that client's own reference data, fixes what it can prove and flags what it can't, and hands the account manager a file ready to submit. Two sides, one source of truth: clients answer only the questions that are theirs, account managers see everything. Owned the whole front end and the checking engine behind it.
Case Study
Problem
A medical billing company preparing insurance claims for its clients did every check by hand. Each client sends its month in its own spreadsheet shape, with its own column names and its own requirements. A specialist read every row, looked up the patient, the provider and the location, and fixed what was wrong before anything went out. A miss surfaced weeks later as a rejected claim, and the knowledge of what each client needed lived in people's heads.
Solution
A portal with two sides. The client side uploads the monthly sheet, sees only the questions it can answer (a patient not on file, a blank cell, an unknown provider) and fixes them in place. The account manager side maps each client's columns once, sets that client's requirements as switches with editable values, runs the check, and downloads either the clean file or a copy with every flag written on the row. Every setting that decides an outcome is visible on screen, and every correction records who changed what and when.
How It Works
The uploaded sheet is read against the client's saved column mapping, with no guessing: an unmapped field blocks the download until a person maps it. Each row is matched to the client's reference data, then checked, and the portal either applies a fix it can prove or raises a flag owned by whoever can answer it. A ledger remembers every line ever exported so nothing goes out twice. Reference-data fixes are stored as an overlay, so the original file stays as evidence. Runs on one Node process with a Postgres store, containerized, with per-user auth and a server-side audit log.
Tools Used
Savings (Time & Cost) — Estimated
Time Saved
Every row of a monthly upload checked automatically instead of read by hand
Monthly
Each client's monthly sheet, previously reviewed line by line by a specialist
Cost Savings
Not measured: the engagement ended before production cutover. The checks target the errors that cause rejected claims, each of which costs a resubmission cycle.
Conservative estimates based on the manual hours each workflow replaces, at $10–25/hr admin rates. Actual savings depend on your volume — figures marked "projected" come from builds validated end-to-end but not yet measured on live client data.
SERVICES
I deliver AI-first automation solutions that eliminate manual work and scale your operations.
Custom Workflows & AI Agents
Design and development of custom automation workflows and AI agents — from business process automation and AI-powered assistants to workflow orchestration, API integrations, and automated content workflows.
Business Process Automation
Streamline operations with Zapier, Make.com, and Go High Level — from lead capture to CRM sync to follow-ups.
AI-Powered Data Pipelines & Integrations
Connect disparate tools and build smart data flows that clean, transform, and route information automatically.
Intelligent Chatbots & Voice Agents
Deploy AI-powered chatbots and voice assistants that handle inquiries, qualify leads, and book appointments 24/7.
Workflow Optimization & Auditing
Audit existing processes, identify bottlenecks, and redesign workflows for maximum efficiency and cost savings.
AI-Assisted Web Development
Landing pages, portfolio sites, and web builds using AI-powered tools. From design to deployment — fast, clean, and production-ready.
Process Documentation & SOPs
Clear SOPs and training materials so your team can run automations confidently — no developer dependency required.
Social Media Automation
Automated content pipelines for LinkedIn, Facebook, and beyond — scheduled posts, AI-generated copy, and engagement workflows on autopilot.
ABOUT ME
I help businesses save time and cut costs by automating operations, training, and support with AI and no-code workflows.
After 3 years building training systems and SOPs at Drivers1st, I now apply that same systems-thinking to create reliable automations — from CRM integrations and lead routing to onboarding and internal workflows.
Tools I work with: n8n, Zapier, Make, OpenAI, CRMs, APIs, and chat automation.
Looking for practical automation that actually works? Let's connect.
WHY WORK WITH ME
Here's what you can expect when we collaborate.
End-to-End Automation
From workflow design to deployment — I handle the full pipeline so you don't have to touch a single manual step.
Measurable ROI
Every automation I build is designed to save time, cut costs, and deliver results you can track from day one.
Fast Turnaround
Most projects are scoped, built, and delivered within 1–2 weeks. Complex systems ship in under a month.
Ongoing Support
I don't disappear after launch. You get documentation and support to keep everything running smoothly.
SKILLS & TOOLS
Core Skills & Expertise
Tech Stack
Automation
AI & Models
CRM & Ops
Google Workspace
Analytics & SEO
Dev & Infra
Comms & Voice
FAQ
Straight answers to the questions I get most.
What kind of automations do you build?+
Business process automations: lead intake and CRM pipelines, AI voice and chat assistants, follow-up sequences, document and quote generation, content scheduling, and internal ops workflows like task routing and stale-deal watchers. If your team does it by hand on a schedule, it can usually be automated.
Which tools do you work with?+
Mainly n8n, Zapier, Make.com, and GoHighLevel for workflows; HubSpot, Notion, Airtable, and Google Workspace on the data side; and AI models like Claude and GPT where language understanding is needed. I work with whatever stack you already have rather than forcing new tools on you.
Do I need AI in my automation?+
Not always — and I'll tell you when you don't. Many of my most reliable builds are deterministic: plain rules, schedules, and triggers with no AI in the loop. AI earns its place when the work involves judgment, language, or unstructured input. Using AI where a simple rule would do just adds cost and unpredictability.
What's the difference between Zapier, Make.com, and n8n?+
Zapier is the easiest to start with and has the most app integrations, but costs grow quickly with volume. Make.com is more visual and cheaper at scale. n8n is open source and self-hostable, which means the most control and the lowest running cost for complex or high-volume workflows — but it needs someone technical to set up. I pick based on your volume, budget, and who maintains it after me.
Can I hire you for a one-off automation, or only ongoing work?+
Both. Plenty of my work is a single, scoped build — one workflow, delivered, documented, done, with no ongoing commitment. Other clients start with one automation and grow into a retainer as more of the business gets automated. We figure out which fits your situation on the discovery call.
How much does an automation project cost?+
It depends on scope, which is why I start with a discovery call: we map the process, I tell you honestly whether automation is worth it, and you get a clear quote before any work starts. Simple single-workflow builds cost much less than multi-system pipelines with AI agents.
How long does a build take?+
Small workflows are usually live within days. Multi-workflow systems — like a full CRM pipeline with AI triage and follow-ups — typically take a few weeks including testing with your real data. You see working pieces early, not a big reveal at the end.
What happens after you deliver?+
You get documentation of how everything works and a walkthrough of the build. Automations include error alerts where it matters, so failures surface instead of silently piling up. I'm available for maintenance and changes as your processes evolve.
GET IN TOUCH
Book a discovery call or reach out directly.
Ted and the live automations are offline
Ted runs on a machine that is down for repairs right now, so he cannot reply. Booking is still fully working, so grab a slot and skip the bot: you will talk to Ed directly.
Book a Discovery Call →The calendar beside this box is live and taking bookings as normal.
Book a Discovery Call
Ted and the live automations are offline
The machine that runs the chatbot and the automation workflows is down for repairs, so Ted cannot reply and workflow submissions will not go through. Booking is unaffected, so the fastest way to reach me is still to grab a call slot.