Skip to content

Capture decision intelligence · GovCon · a HYPATIUS platform

We do not guess.
We engineer the win.

ALIDADE finds the opportunity, scores win probability on seven cited factors, charts the teaming partner on consent, and builds the compliant proposal — on one instrument. Run it in your browser, or call every capability as a governed tool from your AI agent.

24
Tools · six groups
7
Scoring factors, each cited
100%
Of runs written to the ledger
Tool count describes the planned general-availability surface. Ledger coverage is enforced in the data layer, not by policy.
ALIDADE crystalline A mark Bearing dial: months to fielded, by acquisition pathway Three pathways plotted on a 0 to 40 month arc. OTA follow-on, 14 to 19 months, is the recommended bearing. Task order, 10 to 13 months. FAR Part 15, 26 to 34 months. 0 20 40
Months to fielded · by pathway
OTA follow-on 14–19
Task order 10–13
FAR Part 15 26–34
Composite scenario data. No customer information appears.

Two callers, one capability

A console you work in. A tool server your agents call.

Most tools make you choose. ALIDADE runs the same capability both ways, and writes both to the same ledger — so the record does not care which one asked.

▲ The human path

Open the console. Read the bearing.

Eight surfaces — command, bid matches, competitive intel, pipeline, proposals, GovFeed, partners, control room. A capture lead signs the gate. The pursuit moves.

alidade.us → Command
◆ The agent path

Call the same tool. Get the same record.

Twenty-four tools across six groups, exposed over MCP. Your agent scores a pursuit or drafts a section, and the gate still waits for a named person before anything leaves.

tool: score_pursuit → GATE
Both paths write to one run ledger: inputs, tools called, outputs, and the person who signed. Nothing counts until a named person reviews it — which is the same rule, whichever caller asked.
ALIDADE as an MCP tool server: four callers on the left, six tool groups in the middle, the run ledger on the right with AUTO and GATE states.
The tool server. Four callers · six groups · every call on the ledger

The seam

Everything a pursuit needs that is not the writing.

A proposal tool starts when you have already decided to bid. By then the expensive decisions are behind you — whether to bid at all, who fills the gap, which pathway buys the schedule you promised. That is the seam ALIDADE owns. Take a bearing on it before you commit the team.

Who builds it

Producers, primes, subs, and the contracts they hold, from the performer graph.

What it costs

Vendor capacity and pricing intelligence beside the cost model you already trust.

How long to buy

Lead time by pathway, so the schedule includes the time it takes to procure.

Under which authority

OTA, SBIR Phase III, CSO, FAR Part 15 — with the compliance checks each one carries.

Six tool groups · one server

Every surface is also a tool.

The same capabilities your team runs in the console are published to your agent as consent-gated tools. Reads run automatically in-tenant. Anything that acts — contacting a partner, sharing a document, releasing a report — stops at a human gate.

alidade-mcp AUTO · reads GATE · acts
research

Find the opportunity

search_opportunities · enrich_notice · forecast_recompetes · agency_signals

graph

Know the field

performer_lookup · award_history · relationship_paths · incumbent_map

teaming

Fill the gap, on consent

find_partners · fit_score · request_connection · consent_status

logistics

Can it be bought in time

vendor_capacity · procurement_lead_time · pathway_options · sourcing_check

proposal

Write it compliant, cited

compliance_matrix · section_draft · agency_voice · cite_sources

reporting

Prove it

capture_report · pipeline_summary · run_ledger · export_pdf

Tool names describe the planned general-availability surface.
Run ledger · every call, attributable
05:46:01 performer_lookup auto 4 producers · 12 citations
05:46:02 procurement_lead_time auto 3 pathways
05:46:04 request_connection gate Held for approval · J. Broe
05:46:04 export_pdf gate Denied · scope not granted denied
Denials are recorded with the same fidelity as successes. A ledger that only logs what worked is not an audit trail.

The console

Eight surfaces. One instrument.

Screens are product renders on composite exercise data. No customer or operational data appears.

Command surface: the mission queue ranked by score, gates awaiting signature, and pipeline value.
Command. What needs a person today, ranked by score
Bid matches: opportunities scored against a declared capability profile.
Bid matches. Scored against your declared profile
Competitive intel: rivals plotted by incumbency strength against capability overlap.
Competitive intel. Incumbency against overlap
Proposals: the compliance matrix mapped to Section L and M with gaps named.
Proposals. Compliance matrix, gaps named
Control room: the run ledger showing every tool call, its AUTO or GATE state and its authority.
Control room. Every call, its state, its authority

The instrument

One instrument. Every surface an agent or a human works from.

Command briefing, opportunities ranked by probability of win, competitive intel, the Bid Builder with redline and compliance matrix, One Voice, and the agent control room where every tool call is attributable to a person.

Command briefing: the morning bearing on the desktop — scored pursuits, gates awaiting signature, and the run ledger.
Command briefing. The morning bearing, on the desktop.
Opportunities ranked by probability of win, each with its seven factor readings.
Opportunities, ranked by pWin. Seven factors, every one cited.
Competitive intel: rivals in the NAICS plotted by incumbency strength and capability overlap.
Competitive intel. Know the field before the field knows you.
Bid Builder redline view with the compliance matrix mapping Section L to Section M.
Redline and compliance matrix. Sections L and M, mapped and gapped.
One Voice: multiple authors harmonised into a single proposal register.
One Voice. Many authors, one proposal register.
Agent and MCP control room: every tool call attributable to a person and a watermark id, exportable.
Agent & MCP control room. Every row attributable to a person and a watermark id.
Screens are product mockups with composite data. No customer information appears.

In the field

The gate travels.

Three destinations on the phone — Bearings, Ask, Grants. A signature request reaches whoever has to sign it, with the scope it is asking for written out before they approve.

Bearings: pursuits ranked, with a connection request awaiting signature.
Bearings · the morning read
Ask: a plain-language answer with the bearing dial and the tool calls that produced it.
Ask · answer, with its calls
A consent gate: the scope offered, the scope withheld, and the expiry, before approval.
Grants · scope before consent

The scoring instrument

Seven factors. Every one cited.

Seven weighted signals resolve into one probability of win — with the gaps named, so a go/no-go is defensible to the person who has to fund it. The colour is calculated from the score, never chosen. The band carries a mark and a word as well as a hue, so it survives a greyscale print and a colour-deficient reader.

80+
Pursue

Commit the capture team. Gaps, if any, are closeable inside the schedule.

60–79
Consider

Bid only with a partner who closes the named gap. Teaming changes the answer.

<60
Hold

The B&P is better spent elsewhere. The score says why, in writing.

Factor weights · sum 100
FactorWhat it readsWeight
Agency affinityPrior awards, relationships, and incumbency posture at this agency20
Capability matchYour declared capabilities against the stated requirement20
NAICS alignmentCode fit and size-standard position15
Competition levelField density and incumbent strength15
Past performanceRelevance, recency, and CPARS posture10
Certification matchSet-aside eligibility against the solicitation10
Political contextAppropriations, policy shifts, and protest history10
Weights are fixed and published. A score shown in a briefing deck equals the score in the application — the bands and weights are one implementation.

Partner marketplace

No one wins alone.

So connection runs on consent — a signed grant that travels with every tool call, whatever client made it. Discover anonymized partners ranked by bid-match fit. Identities resolve only on mutual opt-in. Nothing your sub's agent asks for leaves your tenant unless you scoped it.

Before consent
Partner · B2F4-9XK

Fills: DoD logistics · OCONUS operations

8(a) · 8+ yrs · ISO 27001

Partner · 7K1C-4QM

Fills: cloud engineering · CMMC L2

HUBZone · 12+ yrs · AWS GovCloud

Handles are watermark identifiers. No firm name, no UEI, no contact until both sides opt in.
The grant

A grant is a signed scope between tenants. Tool groups, fields, and expiry, per grantee. Hidden fields stay hidden to agents too.

SCOPE teaming · proposal
FIELDS capability, past_perf
WITHHELD pricing, staffing
EXPIRES on award or +90d

Many companies. One voice.

When two firms team, two writing styles collide. One Voice harmonizes them into a single compliant narrative while the compliance matrix tracks every §L section against its §M evaluation factor. Forensic watermarking persists on drafts and on final exports.

Governance

Secure by design. Collaborative by architecture.

Contested capture runs on trust. Tenant isolation, an append-only audit record, the consent ledger, and forensic watermarking are not features bolted on for procurement — they are the architecture.

01 · Isolation

Tenant isolation

Row-level security per tenant, enforced in the database. Your data never co-mingles with another prime's.

02 · Record

Append-only audit

Every call recorded with inputs, tools called, outputs, and citations. Denials recorded too. Exportable for the file.

03 · Consent

Scoped grants

Tool groups, fields, and expiry per grantee. Teaming requests reveal identity only on mutual consent.

04 · Authority

Designated approvers

The agent proposes; a person disposes. Read-only tools run automatically. Anything that acts pauses for approval.

05 · Provenance

Forensic watermarking

Persists on drafts and on final exports. Unwatermarked export is gated by tenant policy.

06 · Routing

Provider routing

Model and region routing under tenant control, with security checks in the build pipeline.

Compliance posture · stated plainly

What is true today

  • Database-enforced tenant isolation
  • Append-only audit record, exportable
  • Designated-approver model on every acting tool
  • Tenant-controlled provider routing
  • Security checks in the build pipeline
  • Architecture designed against NIST SP 800-171 controls

What is a pathway, not a status

  • FedRAMP High — target pathway. ALIDADE is not FedRAMP Authorized.
  • NIST SP 800-171 — designed against; no third-party assessment claimed.
  • SOC 2, CMMC, StateRAMP — pending data
We publish the distinction between an architecture and an authorization because your security reviewer will find it anyway. Bring your CISO — we will walk the architecture end to end.

Plans · scale with the mission

Start free. Scale to govern.

From a single capture lead to an isolated, audited enterprise tenancy. Pricing is scoped to your capture shop — enterprise seats are six figures.

Scout
One capture lead, watching the field.
  • Opportunity search
  • Win scoring, read-only
  • Run ledger
Light
A small shop running two or three pursuits.
  • Everything in Scout
  • Pipeline
  • GovFeed signals
Analyst
A capture team with teaming in play.
  • Partner marketplace
  • Scoped grants
  • Competitive intel
Captain
A full capture organization, proposal included.
  • Bid Builder
  • One Voice
  • Compliance matrix
  • MCP server access
Enterprise
Isolated tenancy with governance obligations.
  • Dedicated isolation
  • Gov mode
  • Export controls
  • Named support

Reach us at sales@alidade.us for a tailored quote.

Return-on-investment modelling is available in a briefing, run against your own pursuit history. We do not publish a savings percentage we cannot cite.

STARCHITECT flies the mission.
ALIDADE sources it.

One company · three platforms · hypati.us

The ask

Pick one pursuit. We run it together in 90 days.

You name the capture and the subs. We connect the team, scope the grants, and deliver the cited sourcing section with its ledger.

ALIDADE — Hypatius Capture Intelligence