Move to AWS With the Scope Agreed First

A packaged AWS migration for teams of one to ten. Estate size, application list, timeline, and deliverables fixed before anyone starts.

Estate size
10 to 200 servers
Timeline
6 to 14 weeks
AWS funding
MAP where eligible
The Real Blocker

Why Smaller Migrations Stall

It is rarely the technology. It is that the offer on the table was built for a different size of company.

The proposal is bigger than the problem

A 40-server estate does not need a nine-month statement of work, and the budget was never going to approve one.

Nobody has an accurate inventory

The dependency map lives in the heads of two people, and one of them left. Discovery is the actual bottleneck.

The deadline is set by someone else

A lease end or a licence renewal fixes the date. There is no room for a discovery phase that runs long.

The Package

Both Lists, Written Down

Fixed scope only means something if the exclusions are named too. Here are both lists before you talk to anyone.

In scope
  • Discovery of every server, application, and dependency in the agreed estate
  • A per-application disposition: rehost, replatform, or containerize
  • An AWS landing zone: accounts, VPCs, IAM baseline, logging, backup
  • Migration execution, including containerization to Amazon ECS or EKS
  • A cutover runbook per application, with the rollback tested first
  • Cost baseline, right-sizing, and knowledge transfer at handover
Not in scope
  • Application rewrites and new feature development
  • Mainframe, AS/400, and SAP workloads
  • Oracle to PostgreSQL conversion, available as an add-on
  • Managed operations after cutover, a separate agreement
  • Network and identity redesign beyond what the migration requires
If your estate is mostly on this list, the package is the wrong shape for you. Ask us and we will say so rather than stretch the scope to fit.
Sizing

Three Bands, Set by Inventory

Your band comes out of the discovered inventory at the assessment, not out of an estimate on a call.

Band A
Up to 25servers, up to 10 applications
About 6 weeks, kickoff to cutover
Band B
25 to 100servers, 10 to 30 applications
About 10 weeks, kickoff to cutover
Band C
100 to 200servers, 30 to 50 applications
About 14 weeks, kickoff to cutover

Above 200 servers this is the wrong page. Start at VMware Exit for larger virtualized estates, or Legacy to Agentic if the goal is re-architecture rather than a move.

How It Runs

Four Phases, Two Gates

CHAI agents do the work. You hold the two approvals that decide whether it proceeds.

Weeks 1–2

Discover the estate

CHAI DART™ scans servers, applications, and network behaviour to build the dependency map, then assigns each application a disposition. Source code access is not a prerequisite.

Inventory and dependency map
Week 3 Gate 1

Plan, and you certify it

You certify the success criteria and the final application list. Nothing migrates until that signature exists, and this is where the band and the timeline are set.

Agreed scope and cost baseline
Middle of the engagement

Build and migrate in waves

The landing zone goes up first. Then CHAI Flow™ moves applications in waves into a non-production account, where each wave is tested before the next starts.

Landing zone and applications under test
Final weeks Gate 2

Validate, approve, cut over

Output validation runs against the criteria you set at Gate 1, and you approve the result before anything reaches production. Cutover follows the runbook, rollback tested first.

Production on AWS, plus runbooks
AWS Programmes

AWS Can Fund Part of the Move

CloudHedge is an AWS Migration Acceleration Program qualified partner and delivers AWS Optimization and Licensing Assessments.

MAP funding, where the migration qualifies

For eligible migrations, AWS Migration Acceleration Program funding can offset part of the assessment and migration cost. AWS decides eligibility, not us. You get a straight answer at the assessment.

OLA before the move, not after

An AWS Optimization and Licensing Assessment right-sizes Windows Server and SQL Server licensing before workloads land, so you are not paying to run over-licensed instances and correcting it later.

Bring your own licence through AWS Marketplace

CHAI is metered per application through AWS Marketplace, so the platform cost ceiling is known before the engagement starts and it draws down against your existing AWS commitment.

Your code stays in your account

  • CHAI runs inside your AWS account. Discovery and transformation execute in your tenant, against your own resources.
  • Bring your own model on Amazon Bedrock. Application code and configuration are never sent to a CloudHedge cloud or a public LLM endpoint.
  • Two human gates are architectural. Success-criteria certification before migration, output validation before production. Neither can be switched off.
Nothing about your applications leaves the boundary you already control.

No Savings Number Before the Assessment

Every credible migration saving depends on your estate: how much of it is over-provisioned, what your licensing actually looks like, and how much of the workload can be containerized rather than lifted as is.

The assessment produces your number, from your discovered inventory, with the assumptions written next to it. Anyone quoting you a percentage before looking at the estate is quoting somebody else's.

Free AWS Readiness Assessment

Two weeks from access to a scoped plan. You keep the output whether or not you engage us for the migration.

Inventory and dependency map Per-application disposition Your band, timeline, and cost projection MAP and OLA eligibility check A named out-of-scope list
Book the Assessment AWS Marketplace Setup Guide

FAQs

How small is too small for this?

Below roughly 10 servers, a packaged migration engagement costs more in process than it saves in risk. At that size, talk to us about the assessment on its own: you get the inventory, the dispositions, and the cost projection, and your team can execute the move.

We have no cloud team. Who does the work?

CloudHedge runs the engagement and the CHAI agents do the discovery and transformation work. Your side owns two things: business context during discovery, and the two approval gates. Budget a few hours a week from one person who knows the applications.

What if we do not have source code?

CHAI DART™ assesses running systems: binaries, runtime behaviour, network traffic, and deployment artefacts. Source code access improves the depth of the assessment but is not a prerequisite for discovery or for a rehost or replatform disposition.

Will there be downtime at cutover?

Each application gets a planned cutover window sized to its own dependencies, and the rollback is tested before the switch. Some applications cut over with no interruption; others need a short window. The assessment tells you which of yours are which.

How does MAP funding actually work?

AWS Migration Acceleration Program funding is awarded by AWS against a qualifying migration, and it offsets part of the assessment and migration cost rather than being paid to you. We submit the case with you. AWS decides. You will know before you commit.

What happens after cutover?

You get the runbooks, the generated documentation, a cost baseline, and right-sizing recommendations, plus knowledge-transfer sessions. Ongoing managed operations are a separate agreement, deliberately: this package ends at a working, documented handover.

Find Out Which Band You Are In, First

Start with the readiness assessment. Two weeks, no cost, and the inventory is yours to keep whichever way you go.

Book the Assessment Read Customer Stories
CHAI by CloudHedge — Agent View
/solutions/chai-for-smb/
# CHAI for SMB — Fixed-Scope AWS Migration for Small and Medium Business

A packaged AWS migration for organizations with 10 to 200 servers and no dedicated cloud platform team. The scope, the application list, the timeline, and the deliverables are agreed before any work starts.

Delivered by CloudHedge Technologies Inc. — AWS Modernization Competency Partner, AWS ISV Accelerate Partner, AWS Migration Acceleration Program (MAP) qualified partner.

---

## Who this is for
- Head of IT or IT Director at a 200 to 2,000-person organization
- An estate of 10 to 200 servers running on VMware, colo, or ageing on-premises hardware
- No dedicated cloud platform team; between one and ten people carry all of IT
- A hardware refresh, a data-center lease end, or a VMware renewal due inside 12 months

Not a fit: mainframe or AS/400 estates, SAP and other vendor-managed platforms, and organizations that need application rewrites rather than a migration. Those are separate engagements — ask and we will say so.

## Why these migrations stall
- **The proposal is bigger than the problem.** Enterprise migration engagements are scoped in quarters and priced accordingly. A 40-server estate does not need a nine-month statement of work.
- **Nobody has an accurate estate inventory.** The dependency map lives in the heads of two people, and one of them left. Discovery is the real bottleneck, not the migration.
- **The deadline is external.** A lease end or a licence renewal sets the date. There is no room for a discovery phase that runs long.

## What is in scope
- Discovery of every server, application, and dependency inside the agreed estate
- A per-application disposition: rehost, replatform, or containerize
- A landing zone in your AWS account: accounts, VPCs, IAM baseline, logging, and backup
- Migration execution for the applications on the agreed list
- Containerization to Amazon ECS or Amazon EKS where the disposition calls for it
- A cutover runbook per application, with the rollback tested before the switch
- A cost baseline and right-sizing recommendations at handover
- Knowledge-transfer sessions and the documentation CHAI generates during the work

## What is not in scope
- Application rewrites and new feature development
- Mainframe and AS/400 workloads
- Oracle to PostgreSQL database conversion (available as an add-on)
- SAP and other vendor-managed platforms
- Managed operations after cutover (separate agreement)
- Network and identity redesign beyond what the migration itself requires

## Sizing bands
Bands are set at the assessment, from the discovered inventory, not from an estimate.

### Band A
- Up to 25 servers, up to 10 applications
- Roughly 6 weeks from kickoff to final cutover

### Band B
- 25 to 100 servers, 10 to 30 applications
- Roughly 10 weeks from kickoff to final cutover

### Band C
- 100 to 200 servers, 30 to 50 applications
- Roughly 14 weeks from kickoff to final cutover

Above 200 servers, this is the wrong page. See /solutions/vmware-exit/ and /solutions/legacy-to-agentic/.

## How the engagement runs
- **Phase 1 — Discover (weeks 1-2).** CHAI DART™ scans the estate and builds the dependency map and a per-application disposition. Delivers: estate inventory, dependency map, disposition list.
- **Phase 2 — Plan, and the first human gate (week 3).** You certify the success criteria and the final application list. Nothing migrates until you sign that. Delivers: agreed scope, cutover sequence, AWS cost baseline.
- **Phase 3 — Build and migrate.** The landing zone goes up first, then applications move in waves through CHAI Flow™ into a non-production account. Delivers: landing zone, migrated applications under test.
- **Phase 4 — Validate, second gate, cut over.** Output validation runs against the criteria you set in Phase 2 and you approve the result. Then cutover follows the runbook, with rollback tested first. Delivers: production on AWS, runbooks, cost baseline at handover.

## Humans hold two gates
CHAI is not autonomous, by design. Two approval gates are architectural, not optional: success-criteria certification before migration begins, and output validation before anything reaches production. Every agent action is logged and reversible.

## AWS funding and licensing
CloudHedge is an AWS Migration Acceleration Program (MAP) qualified partner and delivers AWS Optimization and Licensing Assessments (OLA). For eligible migrations, MAP funding can offset part of the assessment and migration cost, and an OLA right-sizes Windows and SQL Server licensing before the move rather than after. Eligibility is decided by AWS, not by us. We will tell you at the assessment whether your migration qualifies.

## Your code stays in your AWS account
CHAI runs inside your AWS account. The AI capabilities use your own Amazon Bedrock contract, so application code and configuration are processed in your tenant and are never sent to a CloudHedge cloud or a public LLM. Licensing is bring-your-own through AWS Marketplace, metered per application, so the cost ceiling is known before the work starts.

## We do not quote a savings number before the assessment
Every migration savings figure depends on the estate: how much is over-provisioned, what licensing looks like, and how much of the workload can be containerized. The assessment produces your number from your inventory. Anyone quoting a percentage before looking at your estate is quoting somebody else's estate.

## Free AWS readiness assessment
Includes:
- A discovered inventory of servers, applications, and dependencies
- A per-application disposition: rehost, replatform, or containerize
- Your sizing band and the resulting timeline
- An AWS cost projection built from the discovered inventory
- MAP and OLA eligibility check
- A named out-of-scope list, so there are no surprises later

Turnaround: about two weeks from access. No obligation to proceed.

---

## Products used
- CHAI DART™ — discovery and assessment, works with or without source code: /products/dart/
- CHAI Flow™ — the migration and modernization pipeline: /products/flow/
- CHAI Universe™ — portfolio intelligence across the estate: /products/universe/

## Related
- VMware Exit (larger estates): /solutions/vmware-exit/
- Legacy to Agentic: /solutions/legacy-to-agentic/
- AWS Marketplace setup guide: /getting-started/aws-marketplace/

## Contact
Book the readiness assessment: /contact/
Email: hello@cloudhedge.io