Self-Hosted AI Agents
Your documents, your customers, your data — none of it leaves your infrastructure.
Every managed AI platform sends your business data to someone else's servers to be processed. Contracts, patient records, financials, customer conversations. We build agents that run on hardware you control — your cloud region, your premises, or fully air-gapped.
Deployed in your account · Full source handover · 7+ yrs building production AI
Should you buy this?
Three reasons to own the stack. Only three.
The law requires it — data-residency rules mean customer or patient data can't be processed abroad
Volume changed the math — past ~100,000 interactions a month, owning costs less than renting
A client contract forbids third-party processing — their compliance obligations are now yours
If none of those is you, don't buy this. Below that volume and without a data rule, the managed track does the same job for a fraction of the cost. Start there, prove it works, and we migrate you when it makes sense — your agents and integrations carry over. Managed track →
Two options
Choose by what your rules actually require.
Everything is identical — same agents, same integrations, same ownership, same handover. One question decides your tier: may an AI provider process your data at all, once identifiers are removed?
SHIELDED
from $8,000
one-time build fee · 8–10 weeks
Your infrastructure, your data. Personal identifiers stripped before anything reaches an AI provider.
Right for
regulated businesses that may use cloud AI providers provided personal data is protected, and high-volume operators who want the economics without running models themselves.
ISOLATED
from $12,000
one-time build fee · 10–14 weeks
Nothing leaves. Not documents, not conversations, not metadata. Ever. Open-source reasoning models run on your own hardware — the agent thinks, decides, and acts entirely inside your perimeter.
Right for
absolute data-residency requirements, air-gapped environments, or any regulator or contract permitting no external processing.
Air-gapped deployment
Shielded
Isolated
Compute requirement
Shielded
Standard compute
Isolated
GPU capacity
Build time
Shielded
8–10 weeks
Isolated
10–14 weeks
Both tiers include the full integration suite
Included in both tiers
At this tier, the full integration suite is included rather than quoted per module.
- Extra CRM & tool integrations
- Additional channels
- Payment processing
- Custom backend & API connectors
- Multi-language support
- Human-in-the-loop approvals
- Custom analytics & reporting
- Monitoring & observability
| Shielded | Isolated | |
|---|---|---|
| Reasoning model runs | Provider API | Your hardware |
| What leaves your perimeter | De-identified text only | Nothing |
| Documents & knowledge base | Your cloud | Your cloud |
| Agent logic & workflows | Your cloud | Your cloud |
| Logs, records, conversations | Your cloud | Your cloud |
| CRM & business systems | Your systems | Your systems |
| PII de-identification layer | Included | Not needed — nothing leaves |
| Third-party outages affect you | Yes | No |
| Model choice | Any frontier provider | Open-weight models |
If you're unsure, you're probably Shielded. We'll say so at scoping rather than sell you the larger build. Isolated exists for requirements that are genuinely absolute.
What the PII shield does on Shielded+
Names, dates of birth, national ID and passport numbers · phone numbers, emails, addresses · card numbers, IBANs, account numbers · medical record and patient identifiers · your own patterns — all tokenised before text leaves your perimeter, and restored locally so the output stays useful.
You also get: a redaction log for every run · detection rules tuned to your vocabulary · a measured accuracy report on your own real data before go-live.
Proof
We've shipped this, not just scoped it.
An on-premise AI document review system for forensic reports — running entirely inside the client's environment, no cloud dependency read the case study →
A fully local voice pipeline — speech recognition, reasoning model and voice, all on the client's own hardware
What we won't claim
Compliance certification.
We build the technical capability that makes data residency achievable — where data flows, where it's stored, what's logged — and hand your compliance specialist the architecture documentation they need. Certifying it against HIPAA, PDPL or GDPR is their work, not ours. Be wary of any vendor who tells you otherwise.
Why it's safe to say yes
Nothing here asks you to commit blind.
Technical discovery comes first, and stands alone — architecture, data-flow map, infrastructure and GPU sizing, deployment plan and cost model. Priced separately, credited in full against the build. The documents are yours whether or not you proceed
See it working before you provision anything — we demonstrate your hardest workflow in our own environment first
Success is a number in the contract — agreed at discovery, tested against your real data
Milestone-based delivery — payments tied to deliverables you can inspect
Revoke our access anytime — we work through a scoped service account that deploys and monitors this system only, never your wider environment. Revoke it and the system keeps running
Deployed into your cloud or hardware, in whatever region you require. We host nothing — you hold the account, the infrastructure and the source
You'll need: a cloud account or hardware in your region · someone who can approve infrastructure provisioning · (Shielded) your own provider accounts · infrastructure budget, sized and quoted at scoping. Isolated needs GPU capacity; Shielded runs on standard compute.
Your security review, procurement process and compliance approvals set the real timeline. We plan around them and tell you at scoping which milestones depend on your side.
Questions
Common questions
How is this priced?+
A fixed build fee plus a monthly retainer for engineering, agreed after scoping. You pay your cloud provider — and, on Shielded, your AI provider — directly, at cost. We never mark up usage.
Are open models good enough for real work?+
For document processing, extraction, classification, support answers and structured decisions — yes, and we'll prove it on your data at discovery. Models are selected against your accuracy, latency and hardware constraints, and we demonstrate real output on your own use case before you commit — not a benchmark chart. For tasks where a frontier model is genuinely required, we'll tell you, and Shielded is the honest answer instead.
Can we start managed and move later?+
Yes — the right sequence for most businesses. Your agents, workflows and integrations carry over. Managed track →
What if you disappear?+
You hold the account, the infrastructure and the full source. Revoke our service account and nothing stops. That's the architecture, not a promise.
Own the stack. Own the data. Own the economics.
Bring your volume, your jurisdiction and your constraints. We'll tell you which tier fits — including when the answer is “neither, yet.”