Platform · Applied AI Infrastructure

Not an LLM wrapper. The layer that makes AI actually useful.

Every Beacon product runs on the same platform: a purpose-built knowledge layer that connects to your existing systems, keeps your documents in your control, and makes every AI query traceable to its source and its cost.

Where your data comes from
ProjectWise OneDrive SharePoint Public Sources
Beacon Knowledge Platform
Connect
reads your existing storage
Index
chunks, embeds, tracks
Route
picks the right model
Cite
links answer to source
Cost
logs every token spent
Where answers surface
TMC-React Estimator Pulse Copilot Any AI interface
No New Systems

It connects to what you already have.

State DOTs have spent decades building ProjectWise repositories, SharePoint document libraries, and public data feeds. Beacon doesn't replace any of that. We read from your existing systems, index the content, and make it answerable.

The document stays where it's always been. The SOP stays in OneDrive. The concept report stays in ProjectWise. Beacon adds a knowledge layer on top — one that keeps in sync as your documents change and answers questions from a single interface instead of five different logins.

The first question agencies ask us is whether this requires a rip-and-replace of their existing infrastructure. The answer is no. We have never asked a DOT to move their documents to get started.

Document & Data Management
Bentley ProjectWise
Engineering workbooks, concept reports, historical projects
OneDrive / SharePoint
SOPs, contact records, policy documents
Google Drive
Shared team documents and agency reports
OpenText / Documentum
Enterprise content management repositories
GIS & Mapping
ESRI ArcGIS
Spatial data, feature layers, ArcGIS Online
HERE Maps
Traffic, road network, and mobility data
Traffic & Operations
RITIS / CARS
Regional incident and travel time data
Skyline SFS / iNet / Claris
Camera feeds, video management systems
Econolite / Yunex / TransCore
ATMS and signal controller platforms
ATSPM / Navigator / 511
Signal performance, traveler info, public feeds
Asset Management & Finance
IBM Maximo / VUEWorks
Asset management and maintenance records
Tyler Technologies / SAP
Financial and ERP systems
AASHTOWare
Project, construction, and maintenance management
Custom / REST API
Agency-specific systems via MCP or direct connection
15-minute sync interval. When a document changes in your storage, the Beacon index reflects it within 15 minutes — automatically, without a manual upload or rebuild process.
How It's Built

Four things every Beacon deployment has to get right.

Trust, Throughput, Tokens, and Transparency. Every product Beacon builds is held to all four.

Trust

Every answer is accountable.

Citations are the floor, not the ceiling. Every response links to its source passage — but trust is built through continuous evaluation. 520 golden Q&A pairs run weekly against every active deployment, scoring retrieval, faithfulness, and citation accuracy against defined thresholds. When something regresses, Beacon staff sees it before operators do.

  • Citations link to the exact source passage, clickable by operators
  • Weekly evals catch hallucinations, broken citations, and stale retrieval
  • Thumbs-down flags route to Beacon staff, not an algorithm
  • Audit log of every query, every answer, every flag
Throughput

The system keeps up with your documents — all of them.

State DOT document environments are large. One state DOT's ProjectWise environment alone contains 10.5 million documents. Beacon is built to index at that scale, across multiple data sources simultaneously, and stay current as documents change. When a file is edited, only the changed section is re-embedded — the rest of the index is untouched.

  • 10.5M+ document environments indexed in production
  • Multiple simultaneous data sources: ProjectWise, OneDrive, ArcGIS, live feeds
  • 15-minute sync — edits surface without manual rebuilds
  • Per-source health monitoring with automated self-probes
Tokens

AI spend is a line item, not a utility bill.

When Uber's CTO admitted he'd burned his entire 2026 AI budget in a few months, the underlying problem was invisible spend. Token spend scales with usage, and usage grows faster than most teams expect. For any organization with a budget owner to answer to, that's not a minor issue — it's a real one.

  • Per-query token cost logged and visible at tokens.beacon.dev
  • Model routing: least-cost model that meets the quality bar for each task
  • Budget caps with auto-downgrade (Opus → Sonnet → Haiku at 95% of cap)
  • Cost-vs-quality dashboard shows the tradeoff before you commit
Transparency

Model liquidity: you can leave any provider, any time.

Beacon is built model-agnostic by design. The knowledge layer — your index, your citations, your evaluation history — lives independently of whichever model is doing the inference. Swap Anthropic for OpenAI or vice versa without rebuilding anything. That's not just a technical convenience; it's leverage. You can credibly take your business elsewhere, which keeps providers accountable to your terms.

  • Model-agnostic routing: any provider, swappable without knowledge loss
  • No cross-agency data sharing — each deployment is isolated
  • Configurable retention: query history matches your records policy
  • Four-role authorization model (Operator / Admin / Beacon Staff / Public)
Data Sovereignty

Your institutional knowledge shouldn't compound into someone else's model.

State DOTs hold decades of irreplaceable operational knowledge — how to price a complex estimate, which responder to call on which corridor, what a pattern of incidents actually means on that stretch of road. Every API call to a frontier model without proper data controls is, at some level, a transfer of that institutional knowledge to a third party whose incentives don't align with yours.

AI model providers have a structural incentive to absorb as much intelligence from their customers as possible — it sharpens their models, which they then lease to your competitors. The question agencies should be asking isn't just "is our data private?" It's "who owns the intelligence our operational data contains, and who does it compound for?"

Beacon's architecture is designed so the answer is always: you. The retrieval index lives within a perimeter your IT team controls. The model sees only the retrieved passage. Your operational knowledge stays yours — and compounds for your agency, not for a model provider.

Your documents stay in your storage ProjectWise, OneDrive, SharePoint — Beacon reads, not copies
The retrieval index is agency-scoped No cross-deployment data sharing; each agency is isolated
Models see retrieved passages only The LLM is never given access to your full corpus
Your data doesn't train the model Beacon operates under Zero Data Retention (ZDR) — prompts aren't stored, aren't trained on, and can't be swept into litigation discovery
No model lock-in — model liquidity by design Swap providers without rebuilding your knowledge layer; your index outlives any single model relationship
Automated Evaluation

You don't have to trust the AI. You have numbers.

Beacon runs a structured evaluation suite on every deployment, every week. 520 golden Q&A pairs — questions with known correct answers — run through the full pipeline and score against six RAG metrics. When something regresses, we know before you do.

The evaluation layer covers the failure modes that actually matter in government AI: hallucinated facts, broken citations, stale sources, and answers that are technically retrieved but not actually relevant. Each metric has a threshold. Drops below the threshold trigger immediate review.

Operators see a summarized trust score. Beacon staff sees the full metric breakdown. When we find a regression, we fix it and tell you what changed — not at the next quarterly review, but within the week.

520 golden Q&A pairs run every week against every active deployment
Keyword Hit Rate
Does retrieval surface the right documents at all?
Content Retrieval
Does semantic search pull the right passage from the right document?
Faithfulness
Does the answer stay within what the retrieved content actually says?
Citation Accuracy
Does the cited source actually contain the answer given?
Relevance
Is the retrieved content actually related to what was asked?
Correctness
Is the final answer factually correct against the known ground truth?
Compliance

Honest about where we are. Clear about where we're going.

Government AI procurement requires vendors to be honest about where they actually stand on compliance — not to oversell certifications they haven't earned. Here's where Beacon actually stands.

We've done the architecture work that makes compliance achievable: threat modeling, NIST SP 800-53 family mapping, a four-role authorization model with explicit no-escalation invariants, and an audit log with hash-chain verification. The foundation is real. The certifications take time.

For agencies with immediate procurement requirements, the path we've used is a public-data-only pilot that requires no DIT or state security review to start — giving agencies six months to work the compliance path in parallel while already seeing value.

NIST SP 800-53 Family Mapping

Architecture mapped against NIST SP 800-53 control families. 24 threat models documented with attacker profiles, mitigations, and residual gaps.

Audit Log with Hash Chain

Every operator interaction logged with tamper-evident hash chain verification and configurable retention to match agency records policies.

GovRAMP / StateRAMP Path

In progress for NCDOT engagement. GCP Assured Workloads architecture scoped. Working through state-specific security review processes.

SOC 2 Type II / FedRAMP

On the roadmap. Not something we'll claim until the audits are done. If your engagement requires it, let's talk timeline — the architecture is built to support it.

Built to Compound

The knowledge flywheel stays inside your agency.

Every query your staff runs generates signal: what was asked, what was retrieved, what was useful. That signal flows back into the knowledge layer — improving retrieval, surfacing gaps in your documentation, and sharpening the system for the next query. Usage → signal → better knowledge → better answers → more usage. That's the flywheel.

That flywheel only compounds value for the organization that owns the index. Run it through an AI model provider's closed system, and the intelligence accrues to them. Run it through Beacon — where your agency owns the index — and it accrues to you. Each product deepens the same substrate: TMC-React indexes your SOPs, Estimator adds historical project data, Pulse brings in 27 public data feeds. Each one makes the next more powerful.

Via a native MCP server, that compounding knowledge is accessible from Microsoft Copilot, Claude, or any AI tool your agency already uses — without custom integration work for each one. The knowledge travels with you — not with the model provider.

Estimator
Adds: 70+ historical workbooks, GIS enrichment layer, task-level cost corpus
TMC-React
Adds: SOPs, contact records, responder route maps, incident logs
Pulse
Adds: 27 public data sources, ATSPM, NAVIGATOR, crash data, live feeds
Beacon Knowledge Platform + MCP
One substrate. Any interface. Your agency's data, answerable from anywhere.
Get in Touch

Want to see the platform in action?

We'll walk you through how Beacon connects to your existing systems and what a deployment actually looks like for your agency.

Start the conversation