Is your software project stuck? I ship it to production.
I step into projects where a software agency, a developer, or wrong technical decisions have led to delays, runaway costs, or code that's hard to evolve. Analysis, recovery plan, execution.
2-4 weeks · Defined scope · Hands-on
Recognize your project in any of these?
- You've already spent tens of thousands of euros, but the product isn't ready.
- The software agency delivers slowly, or with no transparency on what it's doing.
- Every new change breaks something else.
- The tech team (internal or external) has no senior lead making the calls.
- You keep hearing "it's almost done" for months.
- You don't know whether the code is recoverable or should be scrapped.
- You don't know who currently holds the repository, server, domain and database — that is, whether the project is actually yours.
They don't stick for one reason. They stick because they accumulate.
A stack decision taken on the fly. A senior dev leaving. Specs evolving without being documented. Technical debt growing silently. Add 4-5 months and the project seems impossible to finish.
Agencies don't solve these blocks: they have a milestone-delivery model, not continuous problem-solving. Single freelancers often lack the seniority to take over a project in trouble — they can write code, not renegotiate priorities.
Five steps. No magic, no improvisation.
Discovery (3-5 days)
Code access, call with current devs and stakeholders, quick audit of the real state of the project.
Written diagnosis
What's recoverable, what needs rewriting, what to eliminate. Reasoned decisions, not opinions.
Intervention plan
2-4 weeks milestones with measurable acceptance criteria. Before writing code.
Execution
I write code directly or lead your devs. Technical decisions I make and answer for them.
Stabilization
Operational handover to an internal or external dev, documentation. You're not left in a dependency.
Situations where I unblock projects most often.
MVP started 6 months ago, never shipped
Scope reduction to minimum viable, focus on core features, push to go-live in 2-3 weeks.
Stuck legacy migration
Diagnosis of the current situation, strategy decision (big bang vs incremental), push on a concrete direction.
Refactor of main codebase
Clear scope, defined goals, no over-engineering. Verifiable result at every step.
Critical production performance
Diagnosis (profiling, real metrics), targeted interventions with before/after measurement.
Three clear boundaries.
- I don't build products from scratch (that's a separate MVP development).
- I don't manage routine bugs (you need junior devs or an ops team, not a CTO).
- I don't permanently replace your team (that's what fractional CTO is for).
What you ask most often.
I've already paid most of the quote. Where do we start?
Not with the code — with access. The first thing to verify is who currently holds the repository, server, domain, database and third-party accounts: a project whose keys you don't have isn't a late project, it's a project you don't own. It's also the most urgent check there is, because while the vendor relationship is still alive you have someone to ask. After the breakdown it all gets slower and more expensive.
Why have you read the project rather than the vendor?
Because whoever gives you the diagnosis shouldn't have an interest in selling you the cure. I don't resell licences and I have no vendor to place: if the right move for you is staying with your current supplier, or hiring someone other than me, that's what I'll tell you. Sometimes the vendor is right — the spec changed ten times, or decisions never arrived — and I'll tell you that too.
What if the existing code is a disaster?
I say so in the diagnosis. Sometimes the most sensible thing is to rewrite a module, not fix it. I explain why and how much both options cost — you decide.
Can I keep working with my current dev?
Yes, often that's exactly what happens. I enter as the technical reference, the current dev keeps writing code under my guidance. If the current dev is the problem, we address it together in the diagnosis.
What happens after the rescue?
Three options: (1) you take the reins with the product in production, (2) we move to a fractional CTO retainer if you want ongoing reference, (3) we close after handover. Your choice.
How long does a rescue typically last?
2 to 4 weeks of execution, after 3-5 days of discovery. If more time is needed, we're talking partial rebuild, not rescue — and we renegotiate scope/budget at the start, not mid-way.
How much does it cost?
I'll tell you in the first free call, after understanding the situation. Fixed price for defined scope, no open quotes.
Before you rebuild from scratch, read these.
Let's unblock your project.
First a free 30-minute call to understand the situation. Then written diagnosis and a plan in a few days.
Reply within 24 hours · Italy & remote