Project rescue

Someone left you a half-finished system. We'll tell you the truth about it.

The developer stopped answering, the budget is spent, and nobody left in the building can say what the code does or where it runs. Before anyone talks about finishing it, you need an honest read of what you own. That is a five day audit, a flat fee, and a written report that reads the same whether the news is good or bad.

Call (916) 618-0502
5 business days from access granted to a written report you can act on
$1,500 to $3,000 flat audit fee, set by the size of the system before we begin
Credited the audit fee comes off the project if you carry on with us

What is in the 5-day code and infrastructure audit?

The audit from IG Digital Lab runs five business days and ends in a written report: an inventory of every repository and account, a read of the architecture, a security and dependency check, a list of what works and what does not, and cost to finish set against cost to rebuild. The flat fee is $1,500 to $3,000, credited if we continue.

  1. Inventory of repositories and access

    Every repository, branch, server, database, cloud account, domain, DNS zone, third-party service and API key, with a note on who controls each one. Owners regularly discover here that pieces of their system sit in a personal account belonging to somebody who left.

  2. A read of the architecture

    What the system is made of, how the parts talk, where the data lives and what it needs to stay up. You get a plain description and a diagram written for the person paying rather than for a developer, so you can carry it into any conversation that follows.

  3. Security and dependency check

    Exposed keys and passwords in the code, dependencies too old or no longer maintained, missing backups, open ports, and customer data sitting where it should not. We flag severity rather than handing over a list, so you know what has to be fixed this week.

  4. What works and what does not

    We run the system and compare it against what you were told it does: features complete, features half built, and features described in an invoice but absent from the code. This is the part that answers your real question, which is how far along the project is.

  5. Cost to finish against cost to rebuild

    Two numbers with the reasoning behind each, how long each path takes, and what you keep either way. Sunk cost does not enter the recommendation, because money already spent cannot be recovered by spending more on the same code.

  6. The written report

    Everything above in one document, delivered on day five and walked through with you. It is written to be handed to anyone: your board, your accountant, another shop, or the original developer if they resurface.

The fee is flat and quoted before we start, typically $1,500 to $3,000 depending on the size of the system and how many accounts have to be traced. Continue with us and it is credited against the build. Walk away and you still own the report.

Finish, refactor, or rebuild: how we decide

Three possible verdicts with clear criteria for each. We tell you which one before you commit to further work, and we show the reasoning rather than asking you to trust the conclusion.

Verdict one

Finish it

The foundation holds. The remaining work is features, not repair.

  • The architecture fits what the business needs next
  • Dependencies are current, or safely updatable
  • Someone can explain the data model, or the code makes it obvious
  • Cost to finish is well under cost to rebuild
Verdict two

Refactor, then finish

The idea is sound and parts of the build are not. We repair before adding.

  • Core logic is correct but untested and undocumented
  • Security or infrastructure needs work before anything goes live
  • One layer is holding the rest back and can be replaced in isolation
  • Rebuilding would throw away business rules worth keeping
Verdict three

Rebuild

Continuing costs more than starting again. Rare, and we prove it before saying it.

  • Behaviour cannot be verified and nobody knows why it works
  • The platform or framework is out of support with no upgrade path
  • The design cannot carry the scale or the features you need
  • Every change breaks two other things

Whatever the verdict, the work that follows is scoped and priced like any other build, on our normal pricing, and run the way custom software describes. If what stalled was a website rather than an application, see website takeover. If it is offline right now, start with emergency website help.

What you should get handed over

Use this list with any developer, including us. If a project cannot produce these seven things, it is not finished, whatever the invoice says.

  • The repository, in an account your company owns, with full history rather than a zip file of the final state.
  • Hosting and cloud accounts in your company name and on your billing card, with the developer as a member rather than the owner.
  • The domain and DNS at a registrar you can log into, since this is the piece most often stranded in a personal account.
  • Environments, meaning somewhere to test that is not your live system, and a documented way to move a change between them.
  • Secrets and credentials in a password manager you control, not in the code, a chat thread, or one person's head.
  • Documentation covering how to run it, how to deploy it, what depends on what, and what to do when it breaks at 2am.
  • Deployment and backups, meaning a repeatable release process and a restore somebody has actually tested.
Missing most of this list is normal, and it is recoverable. Registrars, hosts and cloud providers all have recovery paths that run through the legal owner of the business, and reconstructing that ownership is part of the audit rather than an extra project.

How we avoid repeating it

Stalled projects are usually a process failure rather than a coding failure. Four habits prevent most of them, and none of them costs extra.

Fixed scope, in milestones

Work is agreed in pieces small enough to deliver and see running. Nothing runs on an open-ended hourly arrangement where the finish line moves every month.

Ownership from the first week

Repository, hosting, domain and credentials are set up in your accounts at the start, so leaving us is always possible and never expensive.

Documentation from day one

Written as the system is built, not promised for the end. That is the difference between a handover and an archaeology project for whoever comes next.

Someone watching afterward

Monitoring, backups, updates and renewals on a plan from $300 per month, so the system does not quietly decay the moment the build ends. See servers and support.

Rescue work is delivered remotely nationwide. Our office is at 3550 Watt Ave, Suite 408, Sacramento, CA 95821, so companies nearby can do the walkthrough in person. New builds are described on our Sacramento software development page.

Start with the audit, not with a proposal

Five business days, a flat fee of $1,500 to $3,000, and a written report on what you own, what it is worth finishing and what that costs. Credited against the project if you continue with us, and yours to take elsewhere if you do not.

See full pricing

Questions owners ask before handing over a stalled project

Can you work with any language or stack?

Most of them. We work day to day in Python, JavaScript and TypeScript, PHP, Go and SQL, on Linux servers, Docker and the usual cloud providers, plus WordPress and the common low-code platforms. If your system is written in something we would be learning on your budget, we say so and tell you what kind of developer to look for instead.

Will you badmouth the previous developer?

No. We assess code, not people. A stalled system usually has a plain explanation: the scope kept moving, the budget ran out, or the working relationship ended. None of that changes what happens next. The report describes what the code does, what it is missing and what it will cost, and it stays readable if you ever show it to the person who wrote it.

What if the code turns out to be unusable?

Then we tell you in writing, with the reasons. Unusable is a real outcome, and week one is a better time to hear it than after another twenty thousand dollars. It usually means behaviour nobody can verify, dependencies too old to update safely, or a design that cannot carry what you need next. We then price the rebuild against what survives, because data, business rules and design work often outlive the code.

How fast can you start?

Usually within a week, and sometimes the same week when something is live and failing. We need read access to the repository, the hosting or cloud account and the domain, plus thirty to sixty minutes with whoever knows the history. When part of that access is missing, recovering it becomes the first piece of work.

Do we have to keep working with you after the audit?

No. The report is yours to take anywhere, including to another development shop or back to the original developer. It is written to be handed over: inventory, architecture, risks, and cost to finish set against cost to rebuild. If you do continue with us, the audit fee is credited against the project.

Find out what you actually own.

Five days, a flat fee, and a written verdict on whether to finish it, fix it or start again.

Or call and text (916) 618-0502. Rescue work is delivered remotely across the United States.