Diagnostic & Repair
Your internal tool breaks every few weeks.
→ DIRECT ANSWER
Internal tools that break repeatedly usually have one of three root causes: original developer no longer involved (no documentation), accumulated patches without architecture (technical debt), or platform integration drift (the substrate changed). The fix is either targeted repair or controlled rebuild — and the right choice depends on the architecture, not the symptoms.
03CAPABILITY LAYER
Why this symptom is a Applications problem.
Most internal tools were built by someone who left, on a platform that's evolved, with patches applied by people who weren't there for the original design. Eventually the patch tax exceeds the value. We diagnose whether to repair, rebuild, or replace — and execute the choice.
See: Applications layerSymptoms
If most of these are true, you have it.
Likely causes
Three patterns that produce most cases.
No system of architecture
Tool was built feature-by-feature without an underlying architecture. Each new patch makes the next one harder. Eventually nobody can predict what changing X will break.
Platform drift
The platform underneath (no-code SaaS, framework, library) has evolved. Original integration patterns are deprecated; the tool depends on undocumented behavior.
Documentation + ownership void
Original builder gone. No tests. No documentation. Operators inherit a black box. Every fix is reverse-engineering before it's engineering.
Recovery process
A 21-day tool stabilization sprint.
Audit + decide
Inventory architecture, dependencies, integration topology. Decide: targeted repair, controlled rebuild, or replacement. Document the case.
Execute the right path
Repair: fix root cause + add tests + document. Rebuild: ship replacement in parallel + cut over. Replacement: migrate to platform that fits.
Verify + hand off
Confirm stability under real workflows. Operator runbook. Monthly maintenance cadence. Optional retainer continuation.
FAQ
Questions we get asked.
01Should we always rebuild instead of repair?+
Almost never. Rebuild is the most expensive option; we recommend it only when repair tax exceeds rebuild cost. Most tools repair correctly with a 1–2 week investment.
02How do you decide repair vs rebuild?+
Repair tax over time vs rebuild cost upfront, weighted by criticality. We model both during the audit so you see the math.
03Can you take over a tool the original developer abandoned?+
Yes — common pattern. We reverse-engineer, document, fix, and either retain or hand back to your team with full context.
04What about no-code platforms (Bubble, Airtable, Retool)?+
We work in all three. Sometimes the right fix is staying on the platform with better architecture; sometimes it's migrating off because the platform hit its ceiling.
05Will this require ongoing engineering?+
For most production internal tools: yes — they need maintenance. We can run that as Embedded Retainer or train your team to own it.
06What does this cost?+
Audit: low five figures. Repair: typical range $10K–$40K. Rebuild: $40K–$150K depending on scope. We model both options during audit.
07How fast can we restore reliability?+
7–21 days for most repair scenarios. Rebuilds take 4–12 weeks but we keep the existing tool running during the transition.
08Can you train our team to maintain it after?+
Yes — that's often the goal. Documentation + runbook + 30-day post-handoff support. Many clients run the tool independently after.
Get it fixed
Stop firefighting the tool. Stabilize it.
A strategy call gets you a clear diagnosis and 21-day stabilization plan within 48 hours.