Notes from the build.
Long-form writing on AI agents, automation patterns, and how I scope, price, and ship production builds — the kind of detail I'd only otherwise share on a call.
Retell AI Multiple Agents on One Phone Number: How It Works
Retell AI now supports multiple agents on one phone number through weighted agent lists and inbound webhooks, but routing needs a clear testing plan.
Retell AI Outbound Calls: Build a Reliable Calling Workflow
Retell AI outbound calls can run follow-up, reminders, and lead reactivation, but the workflow needs clean data, personalization, retries, and compliance controls.
Retell AI Human Transfer: When and How to Escalate Calls
Retell AI human transfer works best when escalation triggers, destination numbers, context handoff, and fallback behavior are designed before the agent goes live.
Retell AI Post-Call Analysis: Track Outcomes Automatically
Retell AI post-call analysis turns call transcripts into structured fields like outcome, urgency, lead quality, appointment status, and next steps.
Retell AI Dynamic Variables: Personalize Every Call
Retell AI dynamic variables let you pass caller-specific context into a voice agent, but they only work reliably when defaults, naming, and test cases are handled upfront.
Retell AI Calendar Booking: Build an AI Appointment Setter
Retell AI calendar booking works when the agent checks real availability, confirms the caller's intent, and writes the booking before promising the slot.
AI Chatbot for Website Conversion
Installing a chatbot and improving conversion are not the same project. Here are the placement, messaging, and testing decisions that determine whether a chatbot actually moves your numbers.
AI Chatbot Pricing for Small Business
Ask five chatbot vendors what it costs and you get five different answers, because "AI chatbot" covers structurally different products. Here is what actually drives the price for a small business.
AI Chatbot After Hours Lead Capture
A visitor at midnight has the same intent as one at 2pm, they just picked a different time. Here is what an after-hours chatbot should handle, and where it needs to be honest about its limits.
Make.com Custom API Authentication
Make's HTTP module can call any API, but it gives you no guardrails around authentication. Here is how to configure API keys, bearer tokens, and OAuth correctly, and debug the 401 when it isn't.
Retell AI Phone Number Setup
Before an agent can answer a single call, it needs a phone number bound to it. Here is how to buy one directly through Retell or preserve an existing number via SIP trunking.
n8n AI Agent Memory
The default memory option that gets an n8n agent working quickly is not the one that should run in production. Here is why Simple Memory breaks under real conditions and what to use instead.
Make.com Webhook Response Guide
Most webhooks never need a custom response, but the moment a caller is actually waiting on a real reply, the default "Accepted" is not enough. Here is how to control it properly.
GoHighLevel Conditional Workflow Logic
Most broken GHL workflow branching comes down to one rule: If/Else evaluates top to bottom and stops at the first match, not the most specific one. Here is how to get the order right.
GoHighLevel AI Employee Guide
GHL's "AI Employee" is really three distinct tools sharing a name and a billing structure. Here is what each one actually automates and where a custom build still makes more sense.
AI Chatbot vs Contact Form
Neither a chatbot nor a contact form is universally better. Here is the honest trade-off, and how to tell which one your business actually needs.
n8n Reverse Proxy Webhook Setup
n8n behind a reverse proxy works fine until you try a webhook and it points to localhost. Here is the specific set of settings that fixes it and the errors that show up when one is missing.
Why Is My Make.com Scenario So Slow?
A scenario that used to take a second now takes fifteen minutes, with no obvious cause. Here is where to actually look before changing anything, and the fixes that matter most.
GoHighLevel Form to Calendar Automation
A form that ends with "we'll be in touch" leaves conversion on the table. Here is how to automate the path from form submission straight to a confirmed calendar booking in GoHighLevel.
GoHighLevel Lead Source Tracking Guide
Most GHL accounts show leads as "direct" regardless of where they actually came from. Here is how native UTM attribution works and where it silently breaks.
AI Chatbot Lead Scoring
A chatbot that treats every conversation identically wastes the most useful thing it produces. Here is how to score leads and route hot ones differently from cold ones.
n8n Retry Failed Node
n8n has retry built into almost every node, but one setting silently disables it without any warning. Here is how retry actually works and the trap that catches nearly everyone.
Make.com API Pagination Guide
A scenario that works fine in testing can silently return only the first page of real data. Here is how offset, page, and cursor pagination actually work in Make.com.
Retell AI Transfer Modes: Cold, Warm & Agentic
Retell supports three distinct transfer modes, plus a separate AI-to-AI handoff mechanism. Here is what actually happens mechanically during each one.
Retell AI Knowledge Base: Setup Guide
Cramming everything into the prompt breaks down fast. A Retell AI knowledge base retrieves the right FAQ or policy mid-call instead, but it introduces its own failure modes worth knowing about.
AI Chatbot Human Handoff
The failure mode that damages trust is not a chatbot needing to hand off, it is failing to recognize that moment. Here is how to design escalation that actually works.
n8n Credentials Security
Every n8n credential is encrypted, but that protection depends entirely on one key. Lose it, and every credential it protected is gone permanently. Here is how to prevent that.
Make.com Iterator vs Array Aggregator
An Iterator splits an array into individual bundles. An Array Aggregator combines them back. Here is when to use each and the source-selection mistake that breaks it.
GoHighLevel Appointment Reminder Workflow
Most appointment reminder workflows break at the same trigger mistake. Here is how to sequence confirmation, reminders, and no-show recovery correctly in GoHighLevel.
How to Build an Inbound Call Workflow in GoHighLevel
Missed call text back covers one outcome. A complete inbound call workflow handles the calls that do get answered too, with tasks, pipeline updates, and routing.
n8n AI Agent Tools
An AI Agent node with no tools connected is just a chatbot in a workflow. Here is how n8n's tool calling actually works and where to add approval gates before it acts.
n8n Sub-Workflow Guide
Copy-pasting the same nodes into every workflow that needs them is how an n8n instance becomes unmaintainable. Here is how sub-workflows solve that properly.
Make.com Router vs Filter
A filter is a single gate. A router splits data into multiple paths. Mixing up the two is one of the most common sources of tangled Make.com scenarios.
How to Integrate Retell AI With a CRM
A voice agent that never touches your CRM is only doing half the job. Here is how to connect Retell AI calls to a CRM without duplicating contacts or losing context.
Retell AI Latency: Causes and Fixes
The difference between a natural-feeling voice agent and an obviously robotic one usually comes down to milliseconds. Here is where that delay actually comes from and what you can fix.
AI Chatbot CRM Integration
A chatbot conversation that never reaches the CRM never really happened as far as your sales process is concerned. Here is how to structure the handoff properly.
n8n Error Trigger Guide
An n8n workflow that fails silently is worse than one that fails loudly. Here is how the Error Trigger node catches failures and routes them somewhere a person will actually see.
How to Fix Make.com Rate Limit 429 Errors
A 429 in Make.com is a pacing problem, not a broken scenario. Here is how Make handles rate limit errors differently from other failures, and how to stop them recurring.
GoHighLevel Round Robin Lead Assignment
Round robin looks simple until you check the weighting, reassignment settings, and whether the user pool still matches your actual team. Here is how GHL round robin really works.
Retell AI Appointment Booking: Routing & CRM Sync
Once the core booking function works, the harder problem is routing the right appointment type to the right calendar and keeping the CRM and reminders in sync with it.
AI Chatbot Appointment Booking
A booking link inside a chatbot is not the same as a chatbot that actually books the appointment. Here is what a real booking flow needs to get right.
n8n Execution Data Cleanup
n8n execution data grows indefinitely unless you turn on pruning. Here is how the built-in retention settings work and what to configure on a production instance.
How to Build GoHighLevel Lead Routing
GoHighLevel lead routing replaces manually deciding who gets the next lead with rules based on source, location, and service. Here is how to build it correctly.
n8n Webhook Authentication Guide
An n8n webhook with no authentication runs for anyone who finds the URL. Here is how to lock it down with the node's built-in Basic Auth, Header Auth, and JWT options.
How Retell AI Function Calling Works
Function calling is what lets a Retell AI agent actually do things mid-call, not just talk. Here is how it works and the mistakes that turn a demo into a flaky production agent.
AI Chatbot Lead Qualification Questions
The technical side of a chatbot is solved. What determines whether it produces real leads is the questions it asks and when it stops asking them.
n8n Queue Mode: Production Scaling Guide
n8n queue mode splits execution across worker processes so a single busy workflow does not block everything else. Here is how it works and when it is worth setting up.
Make.com Webhook Queue Explained
Every Make.com webhook lands in a queue before it is processed. Here is how that queue works, why it backs up, and how to clear or prevent a backlog.
How to Set Up GoHighLevel Missed Call Text Back
GoHighLevel missed call text back turns an unanswered call into an open conversation instead of a lost lead. Here is how to set it up and connect it to a real follow-up sequence.
How to Set Up a Retell AI Webhook
A Retell AI webhook pushes call events to your endpoint in real time. Here is how to set one up, verify it is genuinely from Retell, and connect it to your CRM.
Have a workflow that's burning hours every week?
Bring me one real bottleneck. I'll tell you whether it's worth automating, and what it would take.