SMB AI Adoption Program
AI From Prompt to Process
A 75–90 minute session for small businesses, delivered with community organizations. You bring a process that costs you time. You leave having mapped it, decided where a person must stay in it, and designed a small experiment to find out whether it is worth changing.
It is vendor-neutral, and it is not a product pitch. Some workflows should not use AI at all — the session teaches you how to tell which.
For CPA and professional associations, there is a CPE version: AI in the Back Office
What you leave with
Four artifacts you can use whether or not you ever buy anything from us.
An opportunity map
Where AI may and may not fit in your business — including the places it should not.
One workflow card
Your business maps a single real workflow: its inputs, steps, output, and who reviews what. You keep the card.
A responsible-use checklist
What data may enter a tool, what must not, and what the tool does with it afterwards.
A 30-day experiment
One measure, one baseline, one review date — small enough to actually run.
Map one workflow now
The workflow card is the session's first artifact — and you do not have to wait for a session. Fill it in online, in English or 中文. No name or contact is asked on the card, it is yours to keep, and submitting it contacts no one. At a live session, the paper card is the same card.
Fill in the workflow cardWho it is for
- Owners and operators who can name a recurring process but have no one in-house to design it
- Solo and small professional practices — accountants, attorneys, agents, realtors, consultants — where your own work product is the workflow
- Anyone who has tried AI tools individually and cannot get from "useful answer" to "repeatable work"
How the time is spent
- 0–10 minWhat changed, what did not, and why workflow is the unit that matters
- 10–25 minOpportunity map: the repetitive document, message, research, and handoff work in your business
- 25–45 minA worked example, start to finish — with the human review step visible throughout
- 45–60 minRisk, data, accuracy, access, and vendor boundaries
- 60–75 minBuild your 30-day experiment and decide how you will measure it
- 75–90 minQuestions, and the next step you choose — including none
A deeper follow-up session is designed afterwards from what the room actually brought, rather than delivered from a prepared deck.
Session one
Orlando, September 12, 2026 — hosted by OCPA
The Greater Orlando Chinese Professionals Association ran the first session for its members. The room was mostly people who run something — a delivery operation, a lending business, a software team, the marketing for their own company. Every business mapped one workflow onto a card and left with a 30-day experiment to run on it.
The host published its own account of the morning, in English and Chinese. It is their write-up rather than ours, which makes it the more useful of the two if you are deciding whether to bring a session to your own members.
Upcoming sessions
Sessions are announced together with the organizations hosting them, and registration is handled by the host. Nothing is on the calendar at the moment: a second session and a practice-sharing circle are being designed with the host of session one, and the date will be theirs to announce. If you would like to be told when a session opens near you, send us a note and we will reply when there is a real date and a real host to point you at.
Host a session for your members
We deliver this with chambers, professional associations, colleges, libraries, and business-assistance organizations. The host brings the room and the invitation; we bring the curriculum, the facilitation, and a summary of what the room learned.
What a host gets
- A session built for your members, not a generic deck
- Aggregate learning from the room, back to you
- A follow-up session designed from what the first one surfaced
What a host never gives up
- Your member or attendee list — we do not receive it
- Control of the content to any vendor or sponsor
- Your members' consent: attending is not permission to sell to them
Sponsors can underwrite a session — venue, refreshments, or delivery. Support is disclosed in the room, and a sponsor does not choose the curriculum, the recommendations, or who gets contacted afterwards.
For CPA and professional associations
The CPE version: AI in the Back Office
Designing Human-Gated Workflows for Professional Practices
A hands-on program for CPAs and other professionals who want AI working behind the scenes of their practice without letting it speak for them. Each participant takes one recurring back-office workflow of their own and decides three things: which steps an AI system may draft, which must stay with a person, and which data may never enter a tool. They then work a case from a fictional practice and leave with a 30-day experiment to test the design. Vendor-neutral: no software is sold or taught.
It is built from the session first delivered in Orlando as AI: From Chat to Work (the host's write-up), and adds the case exercise and a short review quiz.
- Designed for 2.0 CPE credits: 100 minutes of content, in a slot of about 110 minutes
- Basic level, with no prerequisites and no advance preparation
- A group program, in person or live over Zoom
- Taught in Mandarin or English
- Recommended field of study: Information Technology
It is offered to associations that issue CPE credit to their own members. The host determines the credit and the field of study, issues the credit, and keeps its own records, including attendance from its own list. For those records, we supply the learning objectives and outline, a bilingual participant evaluation form, the instructor's own evaluation, a facilitator log, and the instructor bio.
The instructor
John Pan is the founder of NESTPILOT LLC, where he builds AI workflow systems for small businesses and professional practices. Before that he spent more than twenty years designing systems in regulated financial services: nineteen years at BNY, most recently as Director, Domain and Cloud Architect, and earlier as an IT architect at JPMorgan Chase and in the Office of the CTO at Nationwide Financial. At BNY he led the architecture of an advisory platform through its move to the cloud, and worked on early AI agents for operations.
He holds an MBA and an MS from The Ohio State University, a BS in physics from Tsinghua University and a BS in computer science from Franklin University, and is a Microsoft Certified Azure AI Engineer and Azure Solutions Architect Expert. A long-time member and former vice president of OCPA, he taught the Orlando session this course is built from.
John Pan 是 NESTPILOT LLC 创始人，为小企业和专业事务所搭建 AI 工作流程。此前二十多年，他在受严格监管的金融服务业做系统架构：在 BNY（纽约梅隆银行）工作十九年，最后担任领域与云架构总监；更早在摩根大通和 Nationwide Financial 首席技术官办公室担任 IT 架构师。在 BNY，他主导了一个投资顾问平台迁移上云的架构，也参与了早期用 AI 智能体运行运维流程的工作。他拥有俄亥俄州立大学 MBA 和理学硕士学位、清华大学物理学学士和 Franklin University 计算机科学学士学位，持有微软 Azure AI 工程师和 Azure 解决方案架构师专家认证。他是大奥兰多华人专业人士协会（OCPA）的老会员和前副会长，本课程所基于的奥兰多工作坊由他主讲。
John Pan on LinkedInAfter the workshop, if you want more
Every next step is optional and opt-in. Attending does not put you on a list. Some businesses run their experiment alone and never speak to us again, which is a perfectly good outcome.
Workshop content is educational. It is not legal, tax, financial, privacy, cybersecurity, or regulatory advice, and it does not guarantee savings, compliance, or business results.