ContactBook free analysis
Secure AI for community banks & credit unions

On-premise AI for community banks & credit unions.

You are examined on your controls. Customer NPI stays inside the perimeter you already defend.

Book a free analysisSee pricing
On-premiseYou own it14-day money-back guarantee
The problem

Your examiner will read your controls, because the rulebook isn’t written yet.

The agencies' 2026 model-risk guidance leaves generative AI out of scope pending a request for information, which means your examiner reads your controls rather than a rulebook. On-premise keeps model weights, inference logs and customer data inside the perimeter you already had validated — so the model validation file examiners look for under SR 26-2 is a document you can actually produce, and data residency is a floor plan rather than a vendor attestation.

Primary sources
The solution

You get a private AI over your own policies, with the model-risk file written as it is installed.

Private policy & procedure Q&A

Front-line staff ask your own policy manual and get the answer with the section cited, instead of guessing or interrupting compliance.

Loan-doc & disclosure summarization

Credit files, disclosures and exception reports summarized for committee, drafted from the documents themselves.

On-prem compliance research

Search your regulatory library, exam responses and prior findings — inside the perimeter, with the inference log kept for validation.

How it works

From analysis to a running system in your office.

01

You tell us where the hours go

A free 30-minute analysis. We say where a private AI pays for itself in your firm — and say so if it doesn’t yet.

02

You approve the build

You see the whole number before anything is ordered. The system is engineered to your workflows, not a template.

03

You get it installed

In your building. Encryption, SSO + MFA and audit logging, documented for your compliance file.

04

You stop thinking about it

24/7 monitoring, model updates and document-pipeline upkeep on one flat monthly plan.

What a build includes

Your policy library and credit files, inside the perimeter you already had examined.

Indexed on hardware in your own control perimeter, answers citing the section, inference logs feeding your log store. The model inventory entry, control narrative and validation file are handed over at install.

See the community banks workload →
written as the system is installed
Validation file
customer NPI crossing the boundary
0
to change your mind, hardware included
14 days
Own it

One price to own it. 14 days to be sure.

One room
$12,000 – $24,000
+ $1,000–$2,500/mo
Whole floor
$32,000 – $56,000
+ $2,500–$5,000/mo
Every floor
Custom
monthly quoted with the build
See the full pricing →
14
days, money back
Every dollar — hardware included. Signed into the build agreement, not a footnote.
Compliance controls

Documented for your compliance file.

Encryption at rest
Full-disk encryption on every drive in the appliance. Keys held by you.
TLS in transit
TLS 1.3 between the chat client and the appliance. Nothing crosses your perimeter.
SSO + MFA
Binds to the identity provider you already run — Entra, Okta or Google Workspace.
Audit logging
Every prompt, document read and answer written to your log store. Retention is your policy.
Data never leaves
No outbound inference calls. Prompts, files, embeddings and outputs stay on the box you own.
Documented for your file
Control narrative, network diagram and configuration handed over as a PDF for your compliance binder.
On-premise is not, by itself, a compliance programKeeping inference in the building removes third-party-disclosure risk — the largest single item. You still need written policy, staff training and a current risk assessment. We hand over the technical control documentation; your counsel or compliance officer owns the program.
FAQ

Community Banks — the questions we get asked.

Can banks use AI under GLBA? →
Can banks use AI?
Yes, and most already do in fraud and credit scoring. Generative AI is newer, so the burden is on your controls: model inventory, validation, change management and an audit trail. Running it on-premise keeps all four inside systems your examiner has already seen.
Which banks are using AI?
The large institutions publish about it; community banks are mostly piloting quietly through core and vendor add-ons. The gap is that those add-ons put customer data in a vendor's cloud, which is the part examiners ask about.
Is customer data safe with AI?
It depends entirely on where inference happens. In a vendor cloud, safety is a contract. On your own hardware, it's a network diagram — customer NPI never crosses the boundary, and you can prove it with logs you own.
What do examiners expect for AI?
A model inventory, a validation file, defined human review, change control and evidence that customer data stayed where your policy says it stays. We hand over the control narrative and configuration documentation for exactly that file.

Thirty minutes, and you’ll know whether this is worth doing.

Sarasota · Bradenton · Lakewood Ranch · Venice · the Suncoast

Book your free 30-minute analysis