Three years stuck. Three months to finish.
That was a CRM migration a national nonprofit had already started twice. I'm Patrick Shea. I do fractional CTO work for companies that have real process pain and nobody in-house whose job it is to fix it.
What this actually is
You get me on a standing schedule, a shared Slack channel, and a straight answer whenever you need one. I own the technical calls you don't have anyone to own right now: what to build, what to buy, what to leave alone, and which vendor is telling you what you want to hear.
Most months are unglamorous. We look at what's slowing your team down, decide what to do about it, and I write down the decision so it doesn't get relitigated in six weeks. When something needs building, I have a team that builds it — scoped and priced on its own, not folded into the retainer.
It's month to month. If it stops being worth what you're paying, you stop paying for it.
Who you'd be working with
Patrick Shea
I've spent about a decade rebuilding the operations of small and mid-sized organizations — CRM migrations that had stalled, custom internal applications for processes no platform covered, integrations between systems nobody designed to talk to each other.
Before Interwoven I was EVP at JASON Learning, a nonprofit founded by the ocean explorer Robert Ballard, where I carried half the annual revenue number and ran the programs and digital media that reached five million students a year. Earlier I was Director of Digital Media at Sea Research Foundation, which operates Mystic Aquarium.
So I've been the person accountable for an operation, not only the person advising one. That's the difference I'd bring to your technical decisions.
A typical first 90 days
So an open-ended retainer doesn't feel open-ended.
Weeks 1–3
Find out what's actually broken
I talk to the people doing the work, not just the people who own the budget. What I'm looking for is the gap between what your software does and what your process needs — the part living in spreadsheets, email threads, and one person's head.
Weeks 4–6
Tell you what's worth doing, in order
A short written list: what to fix first, what it's worth, what I'd leave alone, and what you're being sold that you don't need. Short enough that you'll read it. If the answer is that you don't need me, that's in there too.
Weeks 7–12
Ship the first thing
Something small and real goes into production and gets used. Not a pilot, not a prototype. The point of the first 90 days is that you end them holding working software and a decision you trust, rather than a plan.
What you get
- A standing meeting on a set cadence, weekly or biweekly.
- A shared Slack channel, for the questions that shouldn't wait for the meeting.
- Architecture, build-versus-buy, and vendor decisions — made, written down, and owned.
- Review of the systems and code you already have, including work done by other firms.
- A read on what AI is actually worth in your operation, and where it isn't worth anything.
- Input on technical hiring, and on whether you need to hire at all.
- A team that can build it when advice isn't enough.
What you don't get
- A readiness assessment.
- A maturity model.
- A workshop.
- A forty-page roadmap as the deliverable.
- An offshore team you never meet.
- A recommendation to buy something I have a referral deal with.
Where this fits
Four ways to solve the same problem. This is the honest version of the trade-offs, including where we're the wrong answer.
A full-time CTO
Us
A dev agency
An AI consultancy
When this is the wrong call
Nobody in this category writes this section. It saves us both a call.
You already have a capable engineering leader.
Then you don't need me, you need to back them. A second technical voice at that point creates politics, not progress.
You need a deck for the board more than a working system.
That's real work and it matters. It just isn't this, and there are people who do it better than I would.
You're pre-revenue and looking for a technical cofounder.
You need someone whose upside is your upside. A retainer is the wrong instrument and I'd be taking cash you need.
You want someone to own a problem you haven't decided to fix.
If the organization hasn't agreed the process is broken, no amount of technical help survives contact with it.
One thing about the AI part
Most of what you're being pitched as AI is work we were doing ten years ago, before anyone called it that. It worked then too. AI changed the cost and the speed, not what was possible.
The nonprofit above grew giving 60% — over $6M in new donations — off a CRM migration and the process work around it. No model involved.
The longer version of this argumentWhen it turns into building something
It usually does. That's the honest reason this offer exists: an advisor who spots the problem and has a team that can fix it is more useful than one who hands you a list.
So that it can't become an incentive problem: build work is scoped and priced separately from the retainer, always. You can take the recommendation and give the work to someone else, and people do.
What that's produced
“We tried to get this project right for 3 years, and with you we finished it in 3 months.”
Questions people ask
- How many hours a month is this?
- It varies by what you need, which is why I'd rather talk about cadence than hours. Most engagements are a standing weekly or biweekly meeting plus availability in between. If you need someone forty hours a week, you need to hire.
- We already have a development shop. Does this conflict?
- No, and it's one of the more useful versions of this. A lot of what I do is review work another firm is delivering and tell you whether you're getting what you're paying for. I'll say so plainly either way.
- Who actually does the work? How big are you?
- Three people: two founders and a full-time engineer. That's deliberate and it's the honest answer to the question everyone means when they ask it — the people you meet are the people who do the work. If your project needs thirty engineers, we're the wrong firm and I'll tell you that on the first call.
- What if we want to stop?
- You stop. It's month to month, there's no term, and if it isn't producing value you shouldn't be paying for it. I'd rather end an engagement cleanly than defend one.
- Will you sign an NDA?
- Yes, routinely.
- What does it cost?
- It depends on cadence and how much execution you want included, so I'd rather quote it after one conversation than post a number that's wrong for you. It's a monthly retainer, and I'll give you the figure on the call rather than after a discovery process.
Book an intro call
Thirty minutes. Tell me what's broken and I'll tell you whether I'm the right person to fix it. No deck, no discovery process, and if it isn't a fit I'll say so on the call.
- 1You tell me what's slowing you down, in your words.
- 2I tell you what I'd look at first, and what I wouldn't touch.
- 3If it's a fit, I send you a cadence and a number. If it isn't, I say so.