This document confirms the scope, operating model, and investment for the Interim CTO engagement at Top Notch Products, as discussed on April 20, 2026.
Recommendation: Full-Time cadence. The work breaks into three phases: discovery, build and embed, handoff and continue.
The summary view is the short version. This view is the long one, for the decision-week review.
Work that should be automatic is manual. Knowledge that should be written down lives in a few people's heads. The office move is close, and the business is running at full load while it prepares.
The goal: make the day easier on Kathy and the team. Get more done in less time. Build a foundation that holds when someone is out.
Document the workflows. Automate what can be automated. Train the team on AI, and set up where and how they use it. Build human-in-the-loop checkpoints where automation cannot fit. Cut waste in money, time, and tools, so the team spends more time on what moves the needle.
Starting with Pete and Operations. The work runs the same way each time. I sit with the team. I capture how things actually happen today. I document the workflows in a form that humans and AI can both use. I find quick wins. I scope larger automations and bring them to you for approval before they run. Then we move to the next department.
Quick wins ship as we go. Larger automations queue for your review.
The order is a starting point, not a contract. If you want a different order, we run it that way.
Train employees on using AI. Set up access where it makes sense and not where it does not. Build guardrails and harnesses so the AI runs on real Top Notch context, not generic outputs.
Where automation cannot fit, build human-in-the-loop checkpoints so a person reviews before anything goes out. The point is not to get the AI out of the way. The point is to put the right person at the right checkpoint, with the right context in front of them, and only then let the work move.
Pete signs off on operations changes. Janine signs off on accounting work. Jen signs off on what goes to a customer. The AI gets faster at the drafting; the people stay in charge of the call.
Three cadences. All deliver the same kind of work. The choice is how concentrated the time is. That sets the timeline and the per-day rate.
The opening months are the starting investment. After that we keep going as long as it is valuable. There is no forced end date.
Five categories of value the engagement tends to produce. Bar widths are directional, not committed totals.
These are targets I commit to pursue, not outcomes I can guarantee. The process promises (weekly recap, monthly check-in, 30-day exit) are named in the structure section above.
At SwervePoint I took over an ERP migration already in progress. I had not used either system before: not the one we were moving out of, not the one we were moving into. I pulled reports directly from the database, learned the destination ERP well enough to build my own custom imports and fields, and managed the system myself after go-live. The consultants told Kevin, the owner: "We have never had a client pick up a system and how to customize and administer it as quickly as he did." From that point forward, when other employees had questions, the consultants directed them to me before escalating to Acumatica. That is the same shape of work as understanding Blue Link, building around it, and deciding together if anything ever needs to change at the infrastructure level.
Over four years as Director of IT there: 16 servers, four ERP implementations running side by side, 600+ B2B ecommerce programs for Fortune 500 clients, a 65% cost reduction on the Rackspace-to-AWS migration I designed, and 30+ Fortune 100 security reviews per year. The in-house asset management program I built from scratch saved about $50,000 per year on its own.
Before SwervePoint, seven years at ExecuServices leading 100+ cloud migrations and ecommerce platform moves across Magento, osCommerce, and Drupal Commerce. Twenty years of the same kind of work this engagement asks for, at different scales.
Most weeks I am rebuilding something I already use. The proposal site you are reading is a recent example: the hero video was generated with Gemini and Veo, the site itself is a custom Astro build deployed to a production server. I keep current with the AI tooling because it changes monthly, and the team I bring in alongside me uses what is actually working today.
I live in Danvers, ten minutes from the new office. I run ExecuServices from Beverly.
I do not just advise. I ship. The same patterns power the work I will run at Top Notch.
Adventures in Legal Tech podcast, March 2025. Featured expert on the episode "Site Unseen: How to Build & Maintain a Quality Law Firm Website." (listen)
NECC alumni feature, Fall 2023. Profiled first of four alumni in "NECC Alumni Embrace AI: A Look at the Frontlines of a Booming Industry" (alumNECC magazine). Described as "a natural entrepreneur in the AI innovation realm." (read the feature)
Local. Ten minutes from your new Danvers office.
All three deliver the same kind of work. The choice is how concentrated the time is. That sets the timeline and the per-day rate. The recommendation is Full-Time.
Spreading the work out costs less per month. The runway estimate stretches because every gap between sessions adds re-orienting overhead. Full-Time keeps context warm week to week and gets the most done fastest.
Default plan for Full-Time: three months full time, then three months halftime, going month-to-month so we calibrate as we go. Prefer to lock in six months full time? That option is there.
Halftime and Part-Time each carry a roughly two-year estimated runway to fully optimize. That is a worst-case estimate. We push as much as possible in the shortest amount of time and continue as long as it is valuable. Either side can step the cadence up, step it down, or wind down with 30 days notice.
There is no forced end date on any tier.
Cadence drift.
Gaps between sessions add re-orienting overhead and slow the review cycle. Every gap means one more round of context recovery before meaningful work resumes. Mitigation: weekly written recap to Kathy, monthly explicit re-commit on whether the cadence still fits. If it has drifted, we correct it in writing before the next billing cycle.
Tribal knowledge takes time to capture.
Janine's 25 years of know-how cannot be transferred in a week. The goal is to give her more time for the higher-judgment work that only she can do, not to compress or shortcut the capture process. Mitigation: I sit with her, capture her process through dictation and quick notes, and we shape it into an SOP she owns. Short verification cycles confirm accuracy before anything is locked down. Nothing in her domain gets published without her sign-off.
We work around Blue Link, not against it.
We work around Blue Link where possible and build with it. If something needs to change fundamentally with the underlying infrastructure, that is a bigger conversation. We will scope it together, identify the resources needed, and I will work through the implementation with the team. Any tool that costs more than the current equivalent spend requires Kathy's written sign-off before I act on it. Docuware gets evaluated for rescue-vs.-replace; either path comes to you as a written recommendation first.
Once a cadence is confirmed, the following week is kickoff: