How we deliver

Phased pilots.
Your Git repo. Your control.

You commit one phase at a time, with GO/NO-GO between phases. Your team gets its own Git repository extending ours. The runtime is MIT-licensed, if we disappear, you keep going. Hosting in our Swedish data centre or yours. Built on twenty years of European authority cloud experience, including SÄPO-validated national protected objects.

Phased commitment

Three phases. GO/NO-GO between them.

Most platform engagements ask you to commit to a multi-year, multi-million pilot upfront. We don’t. Each phase is a fixed-price commitment. You decide whether to continue at the end of each. If we miss the mark in Phase 1, you walk away after roughly a third of the total budget, not all of it.

Phase 1

Discovery & demo

4 weeks, fixed price
  • Domain ontology in LinkML
  • One simulated business scenario
  • System-of-action concept demo
  • GO/NO-GO decision documented
GO/NO-GO
Phase 2

Real-data pilot

4 to 8 weeks, fixed price
  • Live integrations (PI/LIMS/ERP)
  • 2 to 3 historical scenarios replayed
  • Audit pipeline live, 5 to 10 daily users
  • System-of-action LIVE
  • GO/NO-GO to production
GO/NO-GO
Phase 3

Production rollout

12 weeks indicative, separate agreement
  • Bidirectional ERP writeback
  • 24/7 monitoring
  • Multi-site rollout
  • Second use case on same platform
Why this matters

The mid-cap European industrial customer cannot easily commit to a multi-year, multi-million pilot upfront, even when the long-term ROI is strong. Phasing solves this. Phase 1 is a structured evaluation at a budget level a sponsor can approve without board involvement. Phase 2 only happens if Phase 1 proved out. The architecture investment compounds: every use case after Phase 2 reuses the same ontology, agents and audit infrastructure.

Your code, your control

Customer-owned repository, MIT-licensed runtime

Vendor lock-in is a real fear. We’ve made it architecturally impossible. Your team gets its own GitHub repository extending our component repos. AEL retains copyright on the runtime; you receive a permanent MIT license. If we disappear, you keep your codebase, your customizations, your data.

01

Your own Git repository

Each customer gets their own GitHub repository extending AEL's open and private component repos. Your customizations, ontology schemas, system prompts, integration configs, custom dashboards, live there. They are your property.

Easy upstream merge by design
02

Peer-reviewed CI/CD

Every PR in AEL's repos triggers automated build + mandatory peer review by the CTO's team before merge. No “silent push” that breaks customer installations. Same discipline applied to large-scale cloud services for European authorities.

No surprise updates
03

Customer-pulled updates

Your repo builds Docker / Kubernetes artifacts. You pull updates via Docker Hub or directly, in your own time. We don't push updates without your control. You decide when a new version lands in production, when to test in staging, when to defer.

Customer-pulled, not vendor-pushed
04

MIT license on the runtime

AEL retains copyright on the runtime (the ontology service, MCP builders, audit pipeline). You receive a permanent MIT license, usage, modification, distribution, sublicensing all permitted. If we disappear tomorrow, you have legal and technical right to keep going.

Continuity built in
Operational maturity

Built on twenty years of European authority cloud experience

We’re not a startup learning operations. Our CTO has spent 20 years operating large-scale cloud services for European authorities, several of which were SÄPO-validated national protected objects. The peer-review, customer-pulled updates, audit pipeline and secure data centre approach all come from that experience.

Background

“Operational maturity from day one, not startup-improvised drift.”

The discipline that runs SÄPO-validated cloud services for European authorities is the same discipline applied to AEL Studio’s CI/CD, peer review, customer update model, and secure data centre architecture in Sweden. Different scale, same principles.

20 years
European authority cloud operations experience
Multiple
SÄPO-validated national protected objects in CTO’s track record
100 %
Peer-reviewed PRs in AEL’s repos before merge
Important nuance

The SÄPO validation refers to systems our CTO operated before AEL Studio. AEL Studio itself is not certified by SÄPO. What transfers is the operational discipline, peer review, customer-pulled updates, secure data centre design, audit-by-default, applied here from day one rather than retrofitted later.

Where AEL Studio runs

You decide where the data lives

Two production-tested options. No middle-cloud, no shared multi-tenant SaaS, no third-party hosting in jurisdictions you can’t audit.

AEL secure data centre, Sweden

Dedicated environment per customer. Own Kubernetes namespaces or dedicated cluster. Own data layer, own access control, your IdP integrated. All traffic and data within Sweden.

Right for you when: you don’t want to manage infrastructure yourself, you want a Swedish jurisdiction guarantee, you need a clear single point of operational responsibility.

Your own infrastructure

On-premise (your data centre), your cloud (Azure / AWS / GCP, your tenant in your region), or hybrid. AEL Studio has no external service dependencies beyond its container images. Air-gapped operation possible.

Right for you when: you have a mature IT organisation, your security policy requires self-managed infrastructure, you have specific air-gap or sovereign cloud requirements.

Start small, decide along the way

Ready to talk about a Phase 1 pilot?

Phase 1 is a 4-week, fixed-price evaluation. You walk away after with a documented domain ontology, a simulated demonstration on a parametrised version of your process, and a structured GO/NO-GO decision, no further commitment.