Air-gapped AI deployment
Runs fully disconnected on your hardware, in your enclave, with open-weight models served locally. No inference call, telemetry beacon, or model-update check ever leaves the boundary — the same platform, minus the internet.
Platform
Enterprise Search One search bar for everything you know Knowledge Graph Every fact carries a receipt Agent Studio AI teammates that follow policy App Builder New One prompt to a live applicationApplications · LACE-GOV
Standards-conformant AI for the mission boundary.
LACE-GOV is the platform packaged for federal, defense, and intelligence programs: BFO/CCO-conformant schema packs, cross-enclave federated discovery without data movement, air-gapped deployment with no external model calls, and accreditation evidence you can export as artifacts.
BFO and CCO ontology conformance
Schema packs align to the Basic Formal Ontology and the Common Core Ontologies, so the knowledge graph you build is interoperable with the semantic standards federal programs increasingly mandate — rather than a bespoke model nobody else can read.
Cross-enclave federated search
Query across enclaves and classification boundaries without moving the underlying data. Each side answers from its own store under its own access rules; only permitted results cross, and every crossing is logged.
Accreditation evidence as an export
Lineage, access enforcement records, and standards-conformance reports come out as artifacts your ATO package can cite — because they are generated by the running system, not reconstructed by hand at assessment time.
Permissions enforced at retrieval
Clearance and need-to-know are evaluated when a document is retrieved, not after a model has already read it. An agent cannot cite what the requesting person is not cleared to see.
RFP and compliance drafting
Grounded drafting over your past-performance library, governing regulations, and prior submissions — each claim traceable to the source document and paragraph it came from.
In practice
An analyst asks a question spanning two enclaves. LACE-GOV federates the query, each enclave answers from its own index under its own access rules, and the response returns with per-fact citations and a record of exactly which sources were reached, by whom, and under what authority — the whole chain available later as accreditation evidence.
Under the hood
Built for the constraints the mission actually has.
No external AI calls, by construction
Air-gapped mode is not a configuration flag over a cloud product. Model serving, embedding, indexing, and orchestration all run inside the boundary, so there is no code path that requires an outbound connection to function.
Provenance that survives audit
Every fact in the knowledge graph is linked to the document, version, and location it was extracted from, with the time it was asserted and the time it was believed true. Point-in-time queries answer what the system knew on a given date — the question investigations actually ask.
Standards packs, not consulting engagements
BFO/CCO alignment ships as versioned schema packs that upgrade with the platform. Conformance is checked continuously rather than certified once and drifting thereafter.
FAQ
What enterprise teams ask.
Can LACE run fully air-gapped with no internet connection?+
Yes. LACE-GOV deploys entirely within your enclave on your hardware, serving open-weight models locally. Inference, embedding, indexing, and orchestration all run inside the boundary, and no component requires an outbound connection — including model updates, which are applied from media you control.
What is BFO/CCO conformance and why does it matter for federal AI?+
The Basic Formal Ontology (BFO) is an ISO/IEC 21838-2 upper ontology, and the Common Core Ontologies (CCO) extend it with mid-level classes for people, organizations, events, and artifacts. Federal programs increasingly require this alignment so knowledge assets remain interoperable across systems and vendors. LACE-GOV ships schema packs aligned to both and checks conformance continuously.
How does cross-enclave federation work without moving data?+
Each enclave runs its own LACE index and answers queries locally under its own access policy. The federating query returns only results the requester is permitted to receive, and every cross-boundary request and response is recorded. The underlying documents never leave their enclave.
What accreditation evidence can LACE-GOV produce?+
Data lineage for every indexed and extracted fact, access-enforcement logs showing which identities retrieved what and under which policy, agent execution traces, and standards-conformance reports. These are exported as artifacts that an ATO or assessment package can reference directly.
Which classification levels and deployment models are supported?+
LACE-GOV is deployed as customer-operated software in the environment you designate — commercial cloud, government cloud, private cloud, on-premises, or fully disconnected. Accreditation is program-specific; the platform supplies the enforcement and evidence, and the authorizing official retains the boundary.
MCP Server
Works where you work.
The whole platform is an MCP server. Connect it once, and the AI tools your team already lives in can search your knowledge, ask the graph, and put governed agents to work — same permissions, same citations, same audit trail, wherever the question is asked.
- Claude Cowork Delegate the busywork
- Claude Code In your terminal
- ChatGPT Ask in chat
- Codex Ships the code
- Microsoft Copilot Across Microsoft 365
- Cursor In your editor
One endpoint — every MCP client
Next step
Bring governed AI inside the boundary.
We set up a pilot on your documents and one workflow your team actually runs. You judge the working output — in weeks, not a quarter-long project.