I'm Patrick Sales — an AI automation specialist working across GoHighLevel, n8n, Make & Zapier. I build AI voice & chat systems, speed-to-lead engines, and end-to-end workflows that turn missed leads into booked appointments, automatically.
Most businesses lose deals in the gaps — a lead that waited 3 hours for a reply, a missed call that never got a text back, a quote that sat in a drafts folder.
I close those gaps. Working across GoHighLevel, n8n, Make, and Zapier, I design automations and AI agents that respond instantly, qualify on their own, book straight into the calendar, and keep the CRM clean — so the team wakes up to booked appointments, not a backlog.
My background is enterprise-grade: 10+ years in automation — from RPA and mainframe at Accenture (American Express, Sempra Energy) to RPA Technical Lead at FPT Software (UNIQLO, Canon, Far Eastern Organization) and Automation Head at GCash. Along the way I've delivered 100+ production bots across enterprise clients — 90+ at Sempra Energy alone at 99% SLA — and 30–50% efficiency gains.
Today that same discipline goes into AI agents and no-code automation — multi-tenant AI front desks on GoHighLevel, Twilio voice agents that book by phone, and 13+ hosted n8n workflows. Systems that actually ship and hold up under real traffic, not demos.
Pick the gap that's costing you the most. Each one ships as a working system wired into your existing tools.
A 24/7 agent that answers calls and webchat, qualifies the lead, checks live availability, and books straight into your calendar — with human escalation when it matters.
Every new lead gets a reply in under a minute across SMS, email, and chat — with round-robin routing, reminders, and no-show follow-ups that keep the calendar full.
Full sub-account setup: pipelines, calendars, A2P registration, workflows, and snapshots — a CRM that's ready to sell from on day one, built to be handed off.
Connect the tools that don't talk to each other. n8n, Make, and Zapier flows that move data, trigger actions, and remove the copy-paste work between your apps.
Retrieval-augmented agents and document pipelines: discovery-call-to-proposal generators, CRM enrichment, and internal copilots that do the reading and drafting for you.
Automated weekly owner reports and always-clean data — dedupe, enrichment, and lead scoring that runs quietly in the background so the numbers you see are the real ones.
A collection of automation systems across GoHighLevel, n8n, Make & Zapier — real builds, deployed and proven, not mockups.
⤢ Full canvas
Pulls listings from three sources into one ranked feed, dedupes them, AI-scores each 0–100 against a skill profile, and drafts an honest cover letter for the winners — posted to Slack for review.
Scanning several job boards by hand is slow, and the best listings are gone before you finish an application.
Scrapes three sources into one canonical feed, dedupes so nothing is seen twice, AI-scores every listing against a skill profile, and drafts a tailored, honest cover letter for the ≥70 winners — never auto-applies.
⤢ Full canvas
Screens every résumé against a data-driven scorecard, posts a Slack review card, and on one human click drafts the interview invite or rejection — then runs onboarding and a weekly HR brief. AI recommends; a human decides.
Screening résumés and running onboarding by hand is slow and inconsistent — and AI hiring tools are regulated, so full automation is a legal risk.
AI scores each candidate on job-relevant evidence only (reasoning logged as an audit trail), a manager approves or declines from Slack, and the right email is drafted for a human to send. Onboarding, a document chaser, and a weekly HR brief run themselves.
⤢ Full canvas
Reads inbound WhatsApp client messages, decides if it's a real request, extracts a structured task, then alerts Slack and files it on a Monday.com board — so no one re-types a request again.
Client requests arrive as casual WhatsApp messages and get lost — someone still has to read each one and re-type it into the project board.
AI filters out chit-chat, extracts a structured task (title, client, priority, budget/deadline) from real requests, alerts the team in Slack, and files it straight onto a Monday.com board via its API.
⤢ Full canvas
New leads trigger an instant, personalized reply in under 60 seconds — logged and tracked across Airtable with Slack alerts.
Leads go cold in minutes. Manual follow-up means most new inquiries never get a fast reply — and the deal goes to whoever answered first.
New leads trigger an instant, personalized response in under 60 seconds, logged and tracked end-to-end across Airtable with Slack alerts.
⤢ Full canvas
Answers 24/7, checks live availability, books into the calendar, routes to the right person, and hands off to a human when it matters.
Missed calls are missed revenue. Reception can't answer every call, book appointments, and triage emergencies at once.
An AI voice agent answers 24/7, checks live availability, books into the calendar, routes to the right person, and hands off to a human when it matters.
⤢ Full canvas
Reads every incoming email, detects intent, drafts an on-brand reply, prioritizes it, and escalates the tricky ones to a human.
Teams burn hours on repetitive customer emails, and slow replies quietly hurt the customer experience.
AI reads every incoming email, detects intent, drafts an accurate on-brand reply, categorizes and prioritizes it, and escalates the tricky ones to a human.
⤢ Full canvas
Detects overdue invoices, sends polite AI-written chasers on schedule, and emails the owner a weekly cash-flow brief — wired to Xero.
Overdue invoices pile up because chasing them is awkward, manual, and always slips down the to-do list.
Automatically detects overdue invoices, sends polite AI-written chasers on schedule, and emails the owner a weekly cash-flow brief — connected straight to Xero.
⤢ Full canvas
Enriches every new contact, scores it against your ICP, removes duplicates, and keeps records clean and sales-ready — automatically.
CRMs rot — missing fields, duplicates, and unqualified leads make the whole pipeline impossible to trust.
AI enriches every new contact, scores it against your ideal-customer profile, removes duplicates, and keeps records clean and sales-ready — automatically.
⤢ Full canvas
Answers from your own docs with grounded, retrieval-augmented responses — and hands off cleanly when it can't help.
Customers ask the same questions around the clock, but support can't answer instantly at 2am.
A retrieval-augmented chatbot answers from your own docs and knowledge base with grounded, accurate responses — and hands off cleanly when it can't help.
⤢ Full canvas
Places outbound calls, holds a natural conversation, qualifies or collects feedback, respects opt-outs, and logs every result.
Calling customers for feedback or qualification doesn't scale — manual dialing eats a rep's whole day.
An AI voice agent places outbound calls, holds a natural conversation, qualifies or collects feedback, respects opt-outs, and logs every result automatically.
⤢ Full canvas
Triages incoming jobs by urgency, matches the best-fit technician, dispatches them, and auto-generates the invoice.
Field-service teams lose time triaging job requests, matching the right tech, and chasing paperwork after the visit.
AI triages incoming jobs by urgency, matches the best-fit technician from the roster, dispatches them, and auto-generates the invoice.
⤢ Full canvas
Crawls competitor pages into a vector database, then answers plain-English questions with grounded, up-to-date intel.
Keeping tabs on competitors means manually checking their sites — and forgetting what changed.
Automatically crawls competitor pages, stores them in a vector database, and lets you ask plain-English questions to get grounded, up-to-date answers.
⤢ Full canvas
An assistant built on the Model Context Protocol that connects to your tools and handles requests from a single chat interface.
Everyday admin — scheduling, lookups, drafting — is scattered across tools and eats focus time.
An AI assistant built on the Model Context Protocol that connects to your tools and handles requests through a single chat interface.
⤢ Full canvas
Drafts on-brand posts, routes them to Slack for one-click approval, publishes to LinkedIn, and logs everything to a sheet.
Staying consistent on social means constantly writing, approving, and scheduling posts nobody has time for.
AI drafts on-brand posts, routes them to Slack for one-click approval, publishes to LinkedIn, and logs everything to a sheet.
⤢ Full canvas
An AI agent that lives in Telegram, understands requests, and responds instantly with the right information or action.
Teams and communities on Telegram need instant answers, but a human can't be on call 24/7.
An AI agent lives in Telegram, understands requests, and responds instantly with the right information or action.
⤢ Full canvas
Reads each inbound WhatsApp lead, scores it against your criteria (BANT), and flags the ones worth a human's time.
Inbound WhatsApp inquiries pile up, and reps waste time on leads that were never a fit.
AI reads each incoming WhatsApp lead, scores it against your criteria (BANT), and flags the ones worth a human's time.
⤢ Full canvas
Reads inbound email leads, qualifies and scores them, and routes the promising ones straight to the right place.
Not every inbound email is a real opportunity, but sorting the good from the noise by hand is slow.
AI reads inbound email leads, qualifies and scores them, and routes the promising ones straight to the right place.
A decade automating for global brands across fintech, retail, utilities & real estate — outcomes over screenshots, since client work stays confidential.
Managed 90+ production bots at 99% SLA across business units through proactive monitoring, incident triage and release readiness — plus Power BI dashboards built from SAP data giving stakeholders visibility into performance, exceptions and trends.
Cut order-processing time 40% while improving data accuracy and operational consistency, and implemented UiPath Document Understanding + AI extraction for intelligent document processing — reducing manual validation effort. Led stakeholder alignment during an onsite New York engagement.
Led sales-process and automation transformation across B2B/B2C operations — running process assessments and target-state design with business leaders, and translating stakeholder needs into scalable automation roadmaps and governance.
Delivered 4 booking and reporting bots across Booking.com and Agoda, saving 50+ hours a week at 98%+ uptime through active production support, issue triage and stabilization.
Supported high-volume transaction environments via batch job monitoring, incident resolution, failure analysis and coordinated fixes — sustaining stable production operations at scale.
Developed finance and reporting automations that improved data accuracy and accelerated recurring operations, with documentation and training materials for adoption and long-term maintainability.
Led cross-functional automation initiatives reducing manual processing 30% and strengthening delivery accountability — standardizing workflows in Power Automate and building Power BI dashboards for progress, risk and utilization visibility.
The platforms I use to design, wire, and ship automation — the ones I reach for on real client work.
What I build, the experience behind it, the stack, and how to start.
Full build teardowns — the architecture, the booking logic, the running costs, and the failures that only show up in production.
The six workflows behind a phone number that answers, checks a real calendar, books the appointment and transfers to a human — plus what it costs to run per month.
Turn on the free feature today — then read where it stops. The build that reads the reply, answers from your own services and hours, books, and escalates.
An honest build-versus-buy comparison — including the cases where you should skip the custom work and go subscribe to a product.
What I build, the stack behind it, how projects run, and what ships with every build.
Rates, availability, timezone overlap — and a five-question checklist for vetting any n8n developer before you hire them.
Tell me where you're losing time or leads. I'll map the automation and show you what it looks like built.