What the engagements are, what they cost, what you own, and where the risks sit. Including the cases where you should hire an agency instead — there are four, and they are further down this page.
Who this is for
A good fit when
- You need a senior engineer to own a hard slice end to end, not to work through a queue someone else wrote.
- You have engineers but nobody senior to check the expensive decisions against.
- Something is in trouble and needs diagnosing before anyone builds more of it.
A bad fit when
- The work is a ticket queue billed by the hour.
- What you need is capacity for years rather than a defined outcome.
- The real problem is scope or organisational, and an engineer is the wrong shape of answer.
The three shapes
Sprint
$4,800per 2 weeks
A focused block for one clear outcome. Good for a rescue, a spike, or a launch push.
- One workstream at a time
- Daily written updates
- Direct Slack access
- Code you own outright
Embedded
$9,200per month
I work as part of your team — standups, reviews, planning. The most common way clients work with me.
- Roughly 3 days a week
- Standups and planning
- Architecture ownership
- Rolling monthly, cancel anytime
Advisory
$1,600per month
For teams with engineers who want a senior second opinion on the calls that are expensive to reverse.
- Weekly call
- Async review
- Architecture decisions
- Hiring input
Which one you need
Two questions settle it. Is the bottleneck capacity or judgement? If your team can build it but keeps making decisions nobody senior has checked, that is Advisory. If the building itself is the constraint, it is Sprint or Embedded. Is the outcome one defined thing, or ongoing? One thing is a Sprint. Ongoing is Embedded, which is what most people end up on.
How an engagement runs
A call, not a pitch
Thirty minutes on what you are actually trying to ship. If I am not the right person, I will say so and usually know who is.
A written scope
Before any money moves, you get the outcome, the approach, the price, and what is explicitly out of scope. No hourly surprises.
Work in the open
Branches you can read, demos over status reports, and a written update every day. You always know where things stand.
A clean handover
Documentation, tests, and a walkthrough with your team. The goal is that you do not need me afterwards.
What I take on
What I turn down
Staff augmentation by the hour
Being a pair of hands on someone else’s ticket queue, with no ownership of the outcome. If the work can be specified as tickets, you do not need my judgement, and you should not pay for it.
Open-ended maintenance retainers
Keep-the-lights-on work with no defined outcome. The goal of every engagement is that you stop needing me; a retainer with no end contradicts it.
Rescues where the problem is not technical
Sometimes a team wants an engineer to fix what is actually a scope problem, or a decision nobody will make. I will say so on the call rather than take the money and confirm the diagnosis six weeks later.
Brand, illustration, native mobile, ML
Outside the discipline. I do interface design and will take a product from rough sketch to shipped, but for these you want a specialist and I will say so early.
The honest comparison
Three ways to get the work done, compared on the dimensions that actually decide it. There is no cost multiple here on purpose — agency pricing varies too much for a number to mean anything, so what is compared is the structure you are paying for.
| Independent engineer | Agency | Permanent hire | |
|---|---|---|---|
| Who writes the code | The person you spoke to on the call. | Often not the person who pitched. The senior name wins the work; the work is done by whoever is free. | The person you interviewed. |
| Ramp-up | Days. | Weeks — a discovery phase, then onboarding several people at once. | Months, after a notice period that is usually one to three. |
| Cost structure | One rate. Fixed scope or a fixed monthly fee, agreed before work starts. | The engineer, plus account management, sales, and the bench between projects. | Salary, benefits, equipment, recruitment fee, and the cost of the months before they are productive. |
| Continuity | Bounded by handover discipline and ownership rather than headcount. One person can still be ill. | Contractual. Bench depth covers absence — the main structural advantage. | Yours, until they resign. |
| Breadth | One deep discipline. Anything outside it needs another specialist. | Many disciplines under one contract and one invoice. | Whatever that one person happens to know. |
| Where the context ends up | With your team, by design — documentation and a walkthrough are part of the work. | With the agency. It leaves when the contract does. | With your team, permanently. |
| Ending it | Two weeks notice. You keep everything built up to that point. | A contract term, often with a minimum commitment. | A redundancy process. |
| Procurement | A sole trader. May not clear enterprise vendor onboarding. | Built for it — insurance, MSAs, certifications. | Not applicable. |
Who writes the code
The person you spoke to on the call.
Often not the person who pitched. The senior name wins the work; the work is done by whoever is free.
The person you interviewed.
Ramp-up
Days.
Weeks — a discovery phase, then onboarding several people at once.
Months, after a notice period that is usually one to three.
Cost structure
One rate. Fixed scope or a fixed monthly fee, agreed before work starts.
The engineer, plus account management, sales, and the bench between projects.
Salary, benefits, equipment, recruitment fee, and the cost of the months before they are productive.
Continuity
Bounded by handover discipline and ownership rather than headcount. One person can still be ill.
Contractual. Bench depth covers absence — the main structural advantage.
Yours, until they resign.
Breadth
One deep discipline. Anything outside it needs another specialist.
Many disciplines under one contract and one invoice.
Whatever that one person happens to know.
Where the context ends up
With your team, by design — documentation and a walkthrough are part of the work.
With the agency. It leaves when the contract does.
With your team, permanently.
Ending it
Two weeks notice. You keep everything built up to that point.
A contract term, often with a minimum commitment.
A redundancy process.
Procurement
A sole trader. May not clear enterprise vendor onboarding.
Built for it — insurance, MSAs, certifications.
Not applicable.
When to hire an agency instead
Four cases where an agency is the better answer. If one of these is yours, I would rather say so on the first call than take the work and have both of us discover it in month two.
You need breadth under one contract
Brand identity, illustration, motion design, native mobile or data science alongside the build, with a single vendor accountable for all of it. Assembling that from independents means you are the one doing the assembling.
You cannot compromise on continuity
The work is business-critical and one person being ill, on holiday, or gone is not a risk you can carry. An agency has bench depth and contractual continuity that a single engineer structurally cannot offer.
The scope genuinely needs a team
Three or more engineers working in parallel against a fixed external deadline. That is an arithmetic problem, not a judgement one, and no amount of seniority solves it.
Procurement requires a vendor, not a person
Liability insurance at a set level, SOC 2, a master services agreement, supplier onboarding built for companies. If your procurement process requires these, an agency is the cleaner route and it is not worth fighting.
The risks, and what bounds them
One person can be ill, on holiday, or unavailable.
Everything is handover-ready throughout rather than produced at the end — documentation, tests and branches you can read, maintained as the work goes. Another engineer can pick it up mid-engagement without an archaeology phase.
The relationship could end badly.
You own the code, the repository and the infrastructure accounts from day one. Nothing is handed over at the end because nothing was ever held.
I could be wrong about something expensive.
A written scope before any money moves, then daily written updates and branches you can read. Being wrong surfaces in days rather than at the end of a quarter.
Capacity is finite. I cannot add people.
True, and worth taking seriously. If the scope needs three engineers in parallel, that is an agency job — see above.
What you own
Code and IP
Yours outright, from the first commit. Not on final payment, not on completion.
Repositories
In your organisation, not mine. I work in yours.
Infrastructure
Your accounts, your billing, your credentials. Nothing runs on mine.
Documentation and tests
Part of the work, never a priced extra.
On handover
A walkthrough with whoever inherits it, live and recorded.
Questions
Still the right fit?
Thirty minutes on what you are trying to ship. If one of the four cases above is yours, I will tell you that instead of pitching.
Currently taking new work — next start october.
Book a call