How is this different from just using AI tools like Claude Code or Copilot?

AI-MSL is not about helping developers code faster — it replaces the need to manage development altogether through an end-to-end AI-powered lifecycle with governance. It operates on a system-wide context (AppGraph) and enforces a governed lifecycle with supervision, ensuring all changes remain coherent, validated, and production-ready.

Why wouldn't I just keep using my existing dev team with AI tools?

You can — but that model still depends on people coordinating, interpreting requirements, reviewing code, and managing releases. AI tools improve individual productivity, not system-level consistency, lifecycle governance, dependency alignment, or long-term maintainability. AI-MSL removes dependency on developer coordination by introducing system-wide context awareness, end-to-end lifecycle execution, and expert supervision at key decision points.

What happens to my code ownership?

You retain full ownership of your code, repositories, and all generated assets. AI-MSL operates on your system in a similar way to a development vendor or internal team — but with full traceability, structured changes, and consistent lifecycle governance. There is no lock-in to proprietary formats or hidden dependencies.

How secure is it to share my repository and data?

AI-MSL follows the same or stricter security model as working with a trusted engineering team or MSP. Your code remains in your repositories, access is controlled and auditable, data is not used to train external models, and outputs are stored in your environment.

What do I actually need to provide to get started?

You need access to your system and the ability to describe your goals. AI-MSL builds system understanding from your existing assets and improves it over time. You don't need perfectly structured documentation — the system evolves its understanding as it works with your codebase.

What do I get after onboarding or assessment?

Even without moving into full development, the assessment stage delivers a structured understanding of your system (AppGraph), visibility into dependencies and risks, identification of modernization opportunities, system quality evaluation, and cost estimates for future maintenance and feature development. This replaces uncertainty with a clear, data-driven baseline.

How do I define what needs to be built?

You define your vision or high-level feature goals. AI-MSL works with you to refine requirements, expand all use cases (not just happy paths), and evaluate impact across the system. Once finalized, development runs automatically through AI agents and is supervised by AI Lifecycle Engineers, who intervene when needed and resolve issues early.

Is this a self-serve platform?

No. While you have full visibility into all stages and can interact with the system directly, AI-MSL is delivered as a platform + managed service. You're supported by an AI Lifecycle Manager who understands your system and goals, and AI Lifecycle Engineers who oversee execution.

Can AI-MSL run in my own VPC or environment?

Yes. Standard packages are delivered through CloudGeometry-managed infrastructure, but enterprise deployments can run in your own VPC or environment, fully configured to your security and compliance requirements.

Is all development reviewed by your engineers?

AI Lifecycle Engineers continuously monitor the development process using quality, correctness, and confidence signals at each stage. When signals are low, they intervene, review outputs, adjust execution, and rerun lifecycle steps. Over time, as the system learns your codebase, accuracy improves and manual intervention decreases.

How is this different from my current development vendor or internal team?

Conceptually similar to having a vendor manage your product lifecycle — but fundamentally different in execution. Requirements are finalized in minutes, not weeks. Development and testing happen in hours. The lifecycle is fully visible and traceable. Execution doesn't depend on specific individuals.

What do I get with the Product Management (PM) package?

A fully structured and validated PRD, including complete requirements, all use cases (including edge cases), and system impact analysis. Generated in hours instead of weeks, and usable with or without continuing into development.

What do I get at the end of the Development package?

A repository with implemented changes, tested and validated code, and updated documentation. All changes are ready to be merged into your main system and deployed.

How can I start safely?

Start with a single application or system component. Run it through AI-MSL, implement a few changes, and compare speed, cost, and quality. Most organizations see clear differences within days.

What if I decide to stop using AI-MSL?

You can exit at any time. You retain your full codebase, improved documentation, structured system understanding, and identified modernization opportunities. Your system will typically be in a cleaner and more maintainable state than before.

What if AI technology changes rapidly — will this become obsolete?

AI-MSL is designed to absorb and adapt to AI advancements, not depend on a single model or tool. We continuously integrate improvements from technologies like Claude, Codex, and others. AI-MSL is not a tool — it's a lifecycle system built on top of evolving AI capabilities.

Will my software become fully automated, self-evolving, and self-healing with AI-MSL?

That's the direction AI-MSL is designed to achieve. The platform enables a closed feedback loop where production signals are turned into improvements, optimizations, and fixes — automatically fed back into the development lifecycle. Human supervision (AI Lifecycle Engineers and Managers) ensures correctness while the system evolves toward increasing levels of autonomy.

How is AI-MSL pricing determined for my system?

Every system is different, so pricing is based on an initial automated assessment that evaluates system complexity, code quality, architecture, and dependencies. Based on this, we estimate the cost of building and maintaining AppGraph, ongoing maintenance and PM support, and expected ranges for future feature development.

Why is maintenance cost different from my current development team cost?

The model is fundamentally different. With a traditional team, you pay for continuous developer capacity, even when no changes are being made. With AI-MSL, you pay a baseline maintenance cost to keep the system understood, monitored, and ready. You don't pay for idle capacity — additional cost is incurred only when new features or changes are implemented.

What is the value of the Product Management (PM) package?

The PM package allows you to go from an idea to a fully structured and complete PRD in hours. AI-MSL expands all use cases (including edge cases), validates requirements against your system, and ensures completeness. The result is a production-ready requirements document that can be used within AI-MSL or by any external team — without additional clarification cycles.

How much will new feature development cost?

Feature cost is determined once requirements are defined — and often estimated even earlier. Pricing depends on feature complexity, system complexity, scope of impact, and expected level of manual supervision. As your system improves and AI-MSL becomes more tuned to it, fewer interventions are needed and costs typically decrease over time.

Plans and pricing

AI-powered development, priced the way AI works

Legacy development sold you time: retained teams, yours or a vendor's. AI-powered development sells outcomes. Always-current system context, precise task-level estimates of time and cost, delivery around the clock, and the pay-per-change model.

See the plans Talk to an expert

Change estimated240 DevCredits, 2 days
PRD approvedby you, before any work starts
Senior engineer sign-offevery change, named owner
Reserved capacityrenews monthly

Trusted by product and enterprise teams

Plans

Three plans.
One governed lifecycle.

Every plan runs the same loop: requirements, an approved PRD, supervised execution, QA evidence and sign-off by a named engineer. The plans differ in where it runs and how much of it is yours.

Company fit:a stable production system in an SMB. The product earns its keep while needs stay steady: security and dependency updates, an occasional feature your clients ask for, and support when something misbehaves, including investigating what happened and why.

MSL One

Best value for SMB

$5K/month
reserved capacity

Start with One

Secure CloudGeometry AWS instance (SaaS)
AppGraph, live system context
PM Portal and requirements finalisation
Time and cost estimate on every change
Software Assessment and Modernisation included
AI-MSL Engineer signs off on every change
Shared AI Lifecycle Manager support
AuditTrail, SSO, RBAC and encryption
Standard support

1 to 2 developersthe monthly task volume they used to handle, in 5,000 DevCredits

Plus pay-per-change for more tasks

Company fit:an actively developed product in a growing company: a living roadmap, monthly releases, parallel workstreams, customers waiting on features, and security questionnaires arriving with the bigger deals.

MSL Pro

Best value for mid-market

$20K/month
reserved capacity

Start with Pro

Everything in One, plus

Dedicated instance on AWS, Azure or GCP, your region
Dedicated AI Lifecycle Manager
Your engineers as supervisors
Continuous modernisation
SAML/SSO, VPN or PrivateLink, IP allowlists, log export
Bring your own model
Security questionnaires answered from process facts and AuditTrail records

4 to 5 person dev teamtheir monthly delivery volume, in 20,000 DevCredits

Plus pay-per-change for more tasks

Company fit:everything stays inside your perimeter, your engineers decide, and every process answers to governance and audit.

MSL Enterprise

For regulated and large organisations

Custom

Talk to an AI Lifecycle Manager

Everything in Pro, plus

On-prem, your cloud or data centre
AppGraph stays in your environment
Your team supervises, with process management
Your IAM, KMS, secrets and SIEM integrated
AuditTrail export to your audit systems
Continuously updated, tested AI stack
24/7 support

Sized to your portfolioannual commitment, enterprise licensing or internal chargeback

Your engineers supervise alongside ours

How you pay

Reserved capacity, not a fee

Pick the plan that matches your needs for development process, system compliance and security. What you pay monthly is reserved development capacity, typically sufficient for your planned workload, and always extendable with instantly estimated changes.

01Monthly DevCredits, replenished
02Estimate before every change
03You approve, then we build
04Renews monthly

Compare plans

What each plan includes

	MSL One $5K/month reserved capacity Start with One	MSL Pro Most popular $20K/month reserved capacity Start with Pro	MSL Enterprise Custom, sized to your portfolio Talk to an AI Lifecycle Manager
Every plan includes governed delivery
Requirements	Included	Included	Included
Approved PRD	Included	Included	Included
Supervised execution	Included	Included	Included
QA evidence	Included	Included	Included
Sign-off	Included	Included	Included
Platform
Hosting	Secure CG AWS instance (SaaS)	Dedicated instance, AWS, Azure or GCP, any region	On-prem, your cloud or data centre
AppGraph, live system context	Included	Included	stays in your environment
PM Portal and requirements finalisation	Included	Included	Included
Time and cost estimate per change	pay-per-change	pay-per-change	Included
Software Assessment	Included	Included	Included
Software Modernisation	Included	Continuous	Continuous
Supervision
Senior engineer sign-off on every change	Included	Included	Included
AI Lifecycle Manager	Shared support	Dedicated	Dedicated, plus your process management
Your engineers as supervisors	—	add one	your team supervises
Security and governance
AuditTrail, chain of custody	Included	Included	Included, plus export to your audit systems
Advanced security	Baseline (SSO, RBAC, encryption)	SAML/SSO, VPN or PrivateLink, IP allowlists, log export	Your IAM, KMS, secrets and SIEM, integrated
Bring your own model	—	Included	plus a continuously updated, tested AI stack
Security questionnaires	On request	Answered from process facts and AuditTrail records	Answered from process facts and AuditTrail records, with your audit team
Support
Support	Standard	Standard	24/7
Pricing
Monthly reserved capacity	$5K 5,000 DevCreditsabout 1 to 2 developers	$20K 20,000 DevCreditsabout a 4 to 5 person team	Custom

Prices in USD. Reserved capacity renews monthly; changes above reserved capacity are billed per authorised change estimate.

Service packages

Optional packages, available on every plan

Add the operating work around the lifecycle when you need it, on top of any plan.

Maintenance

from$2K/month

Proactive scanners and a prioritised maintenance queue, covering security, dependencies and frameworks, handled by MSL.

Talk to an expert

Release Management

from$1K/month

Release readiness, merge and deploy handoff, and the evidence package behind every release.

Talk to an expert

Test Environments

from$2K/month

Integration with your test environment, or we set up a new one and wire validation into it.

Talk to an expert

Deployment and Hosting

from$3K/month

Kubernetes-based Control Plane hosting with 24/7 production support for what MSL builds.

Talk to an expert

FAQ

Questions about the plans

What is a DevCredit?

The unit of a governed, tested, supervised change, delivered under senior engineering oversight. Not raw AI tokens. Every change is estimated in DevCredits before you authorise it.

What happens when I use more than my reserved capacity?

Each change is estimated in DevCredits before you authorise it. Changes above your reserved capacity are billed per authorised change estimate, so you see the cost first and decide. Reserved capacity renews monthly.

Is the monthly amount a subscription fee?

It is reserved development capacity, expressed in DevCredits, sized to the monthly workload a team of that size used to handle. Routine maintenance comes out of it; larger business changes are scoped, estimated and approved on top.

Who does the work, and who checks it?

AI executes within a governed lifecycle; a named AI-MSL Engineer supervises and signs off on every change, and an AI Lifecycle Manager owns delivery. On Pro and Enterprise your own engineers can supervise too.

Who is accountable?

A named AI-MSL Engineer signs off on every change, and an AI Lifecycle Manager owns your delivery. You always know who is responsible, just like a dev team, because it works like one.

Where does my code run?

On MSL One, on a secure CloudGeometry AWS instance. On MSL Pro, on a dedicated instance in the AWS, Azure or Google Cloud region you choose. On MSL Enterprise, inside your own cloud account or data centre.

Your code, your control

Read-only scoped access, encrypted credentials, a full audit trail and self-serve revocation. Your code is not used to train models.

Which plan should I start with?

The Software Assessment, included in every plan, builds the model of your codebase and sets the estimation baseline. Twenty-five minutes with an expert sizes the plan from that, rather than from a guess.

Talk to an expert

Size the plan from your codebase, not a guess.

Bring a roadmap, backlog, vendor, cost or delivery-capacity challenge. In a 30-minute assessment we identify the main constraint, evaluate fit and recommend the plan and a first use case.

A plan recommendation with the reasoning
An estimate before any change starts
You keep control of priorities and final approval

Not running software in production yet, or building a first product from nothing? AI-MSL manages existing software, so this is probably not for you yet.

AI-powered development, priced the way AI works

Three plans.One governed lifecycle.

Reserved capacity, not a fee

01Monthly DevCredits, replenished

02Estimate before every change

03You approve, then we build

04Renews monthly

What each plan includes

Optional packages, available on every plan

Maintenance

Release Management

Test Environments

Deployment and Hosting

Questions about the plans

Size the plan from your codebase, not a guess.

CloudGeometry

Three plans.
One governed lifecycle.