01 — Writing
Notes from the engineering team.
Long-form pieces on what we're building, the operators we serve, and what's actually working in production AI systems. We publish when there is something specific to say.
02 — Posts
Recent writing.
Tool: Missed-call cost estimator — your miss count times your ticket. Not an industry study.
-
2026-08-31
AI receptionist vs hiring a receptionist — what to change in 2026
A hire covers the lobby. An operator covers overflow, after hours, and write-back. Most comparison posts sell a $25 widget. Here is the hybrid that actually works.
-
2026-08-31
AI receptionist vs answering service — which one books the job in 2026
Answering services take messages. Most AI receptionists do the same. Here is how to tell which one actually writes the booking into your system of record.
-
2026-08-25
AI receptionist vs virtual receptionist
A virtual receptionist is a remote human. An AI receptionist is software. The test is whether the booking writes into your system of record.
-
2026-08-21
Why the front desk is the last satisficing system
Every other part of a service business was forced to optimize. The front desk still runs on good enough. That is where the leak lives.
-
2026-08-14
The cost of missed calls for service businesses
How to estimate the leak from your own call logs — not invented industry averages — and what actually closes it.
-
2026-06-02
AI receptionist for med spas — the 2026 guide
Consults, deposits, and Boulevard / Zenoti write-back. What a med spa AI receptionist has to do in production, and where widget deployments fail.
-
2026-06-02
AI receptionist for home services — after-hours and overflow
The 9 p.m. no-cool call is the business. How after-hours voice has to land on the dispatch board, not in a morning voicemail pile.
-
2026-05-17
The boring 80% of production AI
Why most AI agents break in week two: logging, hold-and-merge, transfer rules, and the dashboard a human will actually open.
-
2026-07-27
The Write-Back Thesis — why 60-day AI pilots die
Every “AI receptionist” pilot at a service business that dies at day 60 dies for the same reason: the AI never writes into the system of record. Here is the architecture failure, and how we ship the fix.
-
2026-07-20
Chatbot vs AI operator — what actually ships revenue
Most "AI receptionists" are chatbots with a phone number. Here is what actually makes something an operator, and the three questions that decide whether what you are being sold ships revenue or vanity metrics.
-
2026-07-09
The operational-AI-vs-chatbot line, drawn honestly
A working definition of operational AI, the line that separates it from a chatbot, and the seven questions that decide whether what you are being sold is a system or a wrapper.
-
2026-05-13
What is an AI engineering studio? A working definition
Defines the category that Velzyx operates in, why it is structurally different from an AI agency or a horizontal AI platform, and what an operator should expect from one.
-
2026-05-13
AI engineering studio vs AI agency — a structural comparison
Why most AI agencies are marketing consultancies wrapped in AI vocabulary, and why an engineering studio is a different shape of business.
-
2026-05-13
How to choose an AI partner — a buyer's guide for operators
The questions an operator should ask any AI vendor before signing, the red flags to watch for, and a way to evaluate whether a partner can actually ship.
-
2026-05-13
AI engineering in Newport Beach — why Orange County
Why Southern California is a good place to engineer operational AI, what the Newport Beach engineering scene looks like in 2026, and how Velzyx fits into it.
-
2026-05-13
Why AI demos die in production — and how to engineer systems that don't
An essay on the gap between an AI demo and an AI system. What the demo gets to skip, what production forces you to confront, and what the engineering looks like in between.
-
2026-06-01
AI receptionist for dental practices — what actually works in 2026
An operator's view of AI receptionists for dental practices in 2026. What works in production, where attempts fail, and the architecture that holds up at the practice and DSO level.
-
2026-05-13
Building AI for operational businesses — a field guide
What separates an operation from a consumer product, and why operations need engineered AI configured around the workflow rather than horizontal AI tools.
-
2026-05-13
Why we built Velzyx
A founder essay on the pattern we kept seeing across dental, real estate, and commercial real estate — and the bet on engineered AI per operator instead of horizontal platforms.
-
2026-05-08
How multi-vertical AI engineering wins in 2026
Why specialized engineering per vertical beats horizontal platforms in 2026. Lessons from Toast, Procore, ServiceTitan, and what they teach us about durable AI moats.
-
2026-05-01
Custom-engineered AI vs. off-the-shelf: how to decide
A practical framework for operators weighing custom-engineered systems against off-the-shelf tools. Decision framework, trade-offs, evaluation criteria, and a checklist.
-
2026-04-24
The state of front office automation in 2026
An honest assessment of front office automation in 2026: what's working, what's not, and what comes next as voice AI, scheduling, and async messaging mature.
-
2026-04-17
What we've learned engineering AI for service businesses
Five lessons from engineering our AI voice platform, AgentCentric, and AnalytixCRE: why integration depth beats model sophistication, and why the human handoff is the actual product.
03 — What we publish
The kind of writing we plan to publish.
Subscribe
Want to be the first to read?
No newsletter form yet. Email us and we'll let you know when the next piece publishes.