The answer exists. It's buried in the record.

The right answer is almost always in the record. SphereIQ makes it answerable and auditable — cited to the statute, every action on a signed, immutable ledger.

Book a working session See how governance works

How do government agencies use SphereIQ?

Agencies deploy SphereIQ as a governed Company Brain over statutes, regulations, policy, and case files; agents for records requests, case triage, and procurement checks; and a signed, immutable audit trail of every retrieval and action — deployable self-hosted in your environment, with access governance that inherits source-system permissions. Every answer is cited to the statute, section, and page.

Same platform. Pre-built for your world.

Vertical editions ship with a pre-built Twin schema, agent templates, and compliance packs for the industry — so deployment starts at mile ten, not mile zero.

Company Brain

Statutes, policy & case files

Regulations, program policy, and case histories answerable in seconds — every answer cited to the statute, section, and page, so staff act on the record, not on recollection.

Agent templates

Records, casework, procurement

Public-records (FOIA) request triage and redaction support, case intake and routing, and procurement-document and eligibility checks — governed on every action.

Governance

Built for public accountability

Access scopes that inherit source permissions, a signed and immutable audit trail of every retrieval and action, and self-hosted deployment in your own environment for restricted programs.

Accountable speed. Not one or the other.

Every response cites the statute, section, and page it came from — retrieved from the record, not generated from memory. Every retrieval and governed action is captured in a signed, immutable ledger an oversight body can verify offline. And it runs self-hosted in your environment, where you hold the key that seals stored credentials.

Every answer

Cited to the statute, section, and page it came from

Signed ledger

Every retrieval and governed action captured in an immutable, verifiable audit trail

Self-hosted

Runs in your environment; chat can use a model on your own endpoint

Asked in every evaluation

Can it run in a restricted environment?

It runs self-hosted in your environment, and chat can use a model on a self-hosted endpoint if you choose. It is not air-gapped today: knowledge search calls an OpenAI embedding model, so the deployment needs at least that outbound connection. An air-gapped deployment is on the roadmap; it is not available.

How do we prove what the system did for an oversight review?

Every retrieval and governed action is written to an append-only ledger and can be signed and verified offline — an exportable, tamper-evident record of exactly what the AI did and on whose authority.

Where do agencies usually start?

The highest-volume records or casework queue — or the AI Spend Diagnostic to baseline what’s already running ungoverned across programs before scaling.

The mandate is public. The record should be reachable.

Start with the records or casework queue that carries the most backlog.

Talk to the vertical team Start with a Twin Scan