01
Map
See what is actually there.
Module, service, route, API, dependency, and integration inventory. Dead code candidates. Migration blockers. The map gets cleaner as we read.
Output:A versioned system map with provenance.
Legacy modernization, verification first.
Modernize legacy software without breaking the business.
Grayhaven starts by deciding whether the work should happen. If it should, we map your legacy system, capture critical behavior, build a verification harness, and ship bounded modernization increments with human-reviewed AI assistance.
The first conversation is a fit check, not a commitment to the paid Preflight.
Frontend
AngularJS, Vue 2, jQuery, old React
.NET
Framework, ASP.NET MVC, WebForms, WCF
Java
Spring monoliths, JEE, SOAP services
Why this matters
Old systems do not break because they are old. They break when teams change behavior they cannot see.
AI can write new code fast. The hard part is proving the new code preserves the billing rules, workflow quirks, integrations, reports, and edge cases the business already depends on. A blind rewrite replaces known risk with unknown risk. We do the opposite — map, harness, ship, in that order.
02
Harness
Pin down the behavior.
Critical workflows captured. API contract snapshots. Golden tests for the boundaries that matter. A test that fails loudly is the deliverable, not a green CI run that lied.
Output:A test suite that owns the behavior, not the implementation.
03
Ship
Migrate, incrementally.
Bounded modernization increments sized to one module, route, workflow, or report. Each ships behind a parity check. You can stop at any time and what is shipped continues to work.
Output:Verified increments with rollback notes and evidence.
Fit before scope
The first job is deciding whether we should take the work. Sometimes the useful answer is no.
Credible modernization starts by avoiding work that will not reduce delivery risk. On a fit call, we are looking for the smallest boundary where a Preflight can prove something real.
Good fit
A Preflight makes sense when change risk is real.
- A valuable system is hard to change because behavior lives in old code, workflows, reports, integrations, or edge cases.
- Your team needs evidence before approving a migration, not a rewrite pitch or a staffing plan.
- You can give us repo access, a technical point of contact, and time with the people who know the business rules.
- The right first move is a bounded proof: one boundary, one harness, one verified increment.
Likely not
We should not sell you a Preflight when the premise is wrong.
- You already know exactly what to build and only need extra hands.
- You want a framework rewrite before the current behavior is captured.
- The system is low-stakes enough that manual QA and normal refactoring are cheaper than a harness.
- No one with domain knowledge can validate what the legacy system is supposed to do.
Common modernization patterns
What Grayhaven engagements look like across the legacy lanes we work in.
These examples show the shape and verification posture of the work. They are anonymized composites, not customer case studies.
FrontendPE-backed vertical SaaS
AngularJS dashboard to React, behavior preserved.
Old AngularJS admin dashboard with weak test coverage. Behavior captured against a Playwright harness before any line of React was written.
3 wk
to verified harness
Parity-gated
every shipped increment
AngularJS → React
with parity check
.NETRegulated healthcare software
.NET Framework service stabilized before migration began.
Legacy ASP.NET MVC + WCF surface mapped, dependency risk scored, and API contract snapshots installed in CI before the first service was touched.
All agreed
critical workflows covered
Contract
snapshots in CI
.NET Framework → .NET 8
incremental
RailsMid-market B2B SaaS
Rails monolith made safely changeable.
Test-poor Rails 4 monolith with hidden business rules. Golden-path coverage installed before any service was extracted. Tranches sized to one workflow at a time.
One workflow
per tranche
Rollback
evidence per release
Rails 4 → Rails 7
behind harness
Data & workflowOperations team
Spreadsheet-driven reporting moved into a typed service.
Excel and VBA workflow with embedded business rules. Behavior captured as fixtures, ported to a typed service, and verified with before-and-after output comparison.
Auditable
behavior diff
Fixtures
from production traces
Excel/VBA → typed service
with tests
Java & SpringEnterprise platform team
Spring monolith mapped, harness installed, modules modernized in turn.
Spring 4 monolith with unclear service boundaries. Module-level risk map produced, contract tests added at the boundaries, then modules modernized to Spring Boot 3 one at a time.
One module
per tranche
Verification
package per PR
Spring 4 → Spring Boot 3
incremental
.NETPE operating partner
WCF surface retired behind a REST facade.
Sunsetting WCF service kept callers stable behind a REST facade with contract snapshots. Migration tranches retired the legacy surface one route at a time.
Behavior
captured before change
Bounded
blast radius per tranche
WCF → REST
with snapshots
How a Modernization Preflight runs
Three weeks to the first verified increment for a scoped boundary. Then the same cadence on every tranche after.
- Week 1ReadWe sit with your team and read the code. Stack detection, repo summary, dependency age. No commits, no PRs. Notes only.
- Week 1 to 2MapModule, service, route, and API inventory. Dependency risk. Dead code candidates. Migration blockers. We agree on the boundary of the first increment.
- Week 2Capture behaviorCritical workflows captured. API contract snapshots taken. The safety harness starter goes green on the existing system before anything moves.
- Week 3First proof PROne bounded modernization increment shipped behind a parity check when access, SMEs, and the boundary are in place. Human-reviewed, with rationale, rollback notes, and verification evidence.
- OngoingTranchesSame cadence, smaller blast radius each time. The harness grows. The legacy shrinks. Stop at any time and what is shipped continues to work.
Pricing
Start with a Preflight. Stay on a retainer when the work compounds.
Migration tranches and full programs are scoped separately on the call. You always keep your code and your IP.
Modernization Preflight
Two to three weeks when access, SMEs, and scope are in place.
From
$25,000
- Legacy system map: modules, services, routes, dependencies, blockers.
- Behavior map: critical workflows, API contracts, and side effects.
- Safety harness starter, green on the existing system, ready for CI.
- Migration plan with risk by module and a 30 / 60 / 90 day sequence.
- One bounded first-proof PR when a safe boundary is available, human-reviewed, with rollback notes.
- Executive memo: what can move first, what to leave alone.
Check Preflight fit
Larger or regulated systems are scoped on a call. If the first proof should be a work order instead of a PR, we will say so before the Preflight starts.
Maintenance Retainer
Ongoing. Monthly cadence, stop any time.
From
$15k/mo
- Test and harness maintenance as the codebase evolves.
- Monthly dependency risk and security review.
- Continued retirement of legacy surfaces on a known cadence.
- Additional migration work orders, scoped and verified.
- Code quality and architecture reviews against the harness.
- Production triage for issues that touch modernization work.
Discuss retainer fit
Designed for systems already under harness. You keep your code and your IP; we keep the safe-change layer healthy.
Start small.
Book a modernization diagnostic. Leave with a first boundary, a risk map, and a recommended next step.
We are deliberately slow to take on engagements. If we are not the right fit, we will say so on the first call.
Contact
Ask whether this system is worth a Preflight.
Show us the system your team cannot safely change. We’ll tell you whether a Preflight makes sense, what boundary we would inspect first, and whether a first verified increment is realistic with the access and domain knowledge available. If we’re not the right team, we’ll say so.
Location
United States
Phone
(405) 320-8212
Hours
Monday - Friday
8:00 AM - 7:00 PM
8:00 AM - 7:00 PM