Three ways to engage. Each one stands alone.
Most customers move sequentially: Audit, then Blueprint, then Rebuild, then Manage. Some only need one or two. You're never locked into the next tier, and the audit is yours to keep either way.
Audit + Blueprint
Understand what you have. Decide what's worth rebuilding.
Free → $25K
one-time • 1–3 weeks
The Audit is free. It produces a Digital Twin of your product: every service, every feature, every dependency, every hot spot, mapped, scored, and ranked by what's slowing you down and what's worth carrying forward.
The Blueprint turns the Audit into a decision document: which features to keep, which to simplify, which to drop, what stack to land on, what order to rebuild in, what it will cost, and what it will take. Your team approves it before a line of new code is written.
What you get
- Digital Twin of your existing product: full dependency map, feature inventory, usage and revenue attribution, risk and tech-debt scoring
- AI-readiness assessment: what's blocking you from shipping modern AI features today
- Feature triage: keep, simplify, or drop, with the data behind each call
- Target architecture proposed on the stack of your choice
- Phased rebuild plan with dates, scope, and budget
- One-week turnaround on the Audit. Two more weeks for the Blueprint.
You walk away with
A decision-ready plan you can take to your board, your CFO, or your internal team, whether or not you continue with us.
Rebuild
Ship the new product. We do the work. Your team reviews.
$100K – $1M
outcome-based per engagement • 2–6 months typical
This is where most of the value is. Specialized agents rebuild your product on the modern stack agreed in the Blueprint (services, schema, APIs, frontend, tests) in parallel, around the clock. Old and new run side-by-side until cutover. Your team reviews every PR through your existing process.
What you get
- Production-ready codebase on your chosen modern stack
- Feature parity with the proven features from your legacy product, lean and idiomatic on the new stack
- Test coverage generated alongside the code, not bolted on after
- Database migration tooling and reversible cutover plan
- Parallel-run mode: old and new live simultaneously, traffic shifted gradually
- Full handoff documentation, architecture decision records, and runbooks
- Every PR goes through your existing review process. No code lands without your team's approval.
You walk away with
A modern product running in production, owned by your team, with no Merven dependencies you can't walk away from.
Why outcome-based pricing: You pay for what gets merged and shipped, not for hours logged. If a rebuild takes us twice as long as planned, that's our problem. If we deliver in half the time, you get the benefit.
Manage
Keep it modern. Forever.
$8K – $40K
per month recurring • month-to-month
Once the new product is live, agents stay on to keep it from aging. Dependency upgrades, security patches, framework migrations, performance regressions, deprecation tracking: the maintenance work that quietly compounds into the next rewrite, handled continuously.
What you get
- Continuous dependency and security patching
- Framework version upgrades as they ship (React 19, Node LTS rolls, Postgres majors, etc.)
- Performance regression detection and remediation
- Compliance posture monitoring (SOC 2, HIPAA, GDPR, DORA, depending on your needs)
- Quarterly architecture reviews: proactive recommendations before debt accumulates
- Same review-and-approve flow as Rebuild. Nothing ships without your team's sign-off.
You walk away with
A product that's still modern in three years. Not because your team did the work, but because it got done while they were shipping features.
Why month-to-month: No long-term commitment. If we stop earning the budget, you stop paying.
Inside the Engagement
How a Rebuild actually runs.
The Rebuild tier is where most of the work happens. Here's what the six weeks after Blueprint approval look like in practice.
Connect
We integrate with your repo, CI, observability, and ticket system. Read-only at first. Your security team approves every access scope.
Mirror
Agents stand up the new stack in a parallel environment. Same data shape, different foundation. Nothing in production touched.
Rebuild in slices
Features get rebuilt one logical slice at a time: auth, billing, core workflows, dashboards. Each slice is independently testable and independently shippable.
Review
Every PR lands in your repo, in your review process. Your engineers approve. Your CI runs. No backdoor merges.
Parallel run
Traffic gets routed gradually: 1%, 10%, 50%, 100%. Both stacks live. Real customers hit the new stack while the old one stands by. Instant rollback at every step.
Cutover and decommission
When you're confident, old stack goes offline. We hand over runbooks, architecture docs, and on-call materials. Your team owns it from here.
Stacks
Stack-fluent, by design.
We rebuild on the stack your team can hire for, operate, and ship on, not the one we're locked into. The list below is where we have the deepest fluency; we work outside it on request.
Languages
TypeScript, JavaScript, Python, Go, Java, Kotlin, Ruby, C#, PHP, Rust
Frontend
React, Next.js, Vue, Svelte, Remix, Astro
Backend
Node, FastAPI, Django, Rails, Spring, .NET
Data
Postgres, MySQL, MongoDB, DynamoDB, Snowflake, BigQuery, Redis, Kafka
Infra
AWS, GCP, Azure, Kubernetes, Terraform, Pulumi
Legacy we migrate from
PHP 5/7, AngularJS, Rails 3/4, jQuery monoliths, .NET Framework, older Java EE, custom ETL pipelines
If your stack isn't here, ask. We've probably seen it.
What Merven Isn't
A few things we don't do, on purpose.
We don't do greenfield from a spec.
We rebuild what you already have, leaner. If you're starting from zero, you don't need us.
We don't do lift-and-shift.
Copying a bad codebase onto a new framework just buys you the same problems on a faster runtime. We rebuild deliberately, feature by feature.
We don't sell a platform.
There's no Merven SaaS your team has to log into and learn. The artifact is your codebase. The interface is your repo.
We don't lock you in.
No proprietary frameworks, no Merven-only runtimes, no clauses that make leaving hard. Everything we ship is standard, idiomatic code on the stack you chose.
We don't replace your engineers.
Your team approves every change. The goal is to give them a codebase they're excited to own, not to make them redundant.
Compare
Merven vs. the alternatives.
| Merven | Internal rebuild | Big consultancy | |
|---|---|---|---|
| Time to cutover | 2–6 months | 12–24 months | 9–18 months |
| Cost | $150K–$1M, fixed | $1.5M–$5M+ in salaries, opportunity cost | $2M–$8M, hourly |
| Roadmap impact | None. Your team keeps shipping | Severe. Best engineers pulled | Moderate. Your team supervises |
| Stack choice | Yours | Yours | Often theirs |
| Code ownership | Yours, fully | Yours | Yours, sometimes with strings |
| Post-launch care | Optional Manage tier | Your team again | Hourly, indefinitely |
| Risk of failed cutover | Low. Parallel run, reversible | High. Pause-and-pray | Medium. Handoff problems |
FAQ
Solutions, in detail.
See your rewrite, costed and scheduled, before you commit to anything.
Free Audit. One week. You walk away with a Digital Twin of your product, a feature-triage proposal, a target architecture, and a real timeline with a real number. Whether you build with us, build internally, or don't build at all, the audit is yours to keep.