The procurement surface Folio Bid reads.

01. What the engine does

Folio Bid sits where the capture lead sits. Its job is to find every Canadian tender that could credibly fit the firm, compile the compliance matrix that decides the bid before narrative is read, and assemble the response volumes from a structured clause library under a compressed clock. The engine is an AI-assisted workbench, not an autonomous agent. Senior capture leadership signs every submission before it leaves the system.

The stack is built on two named engines. Harbour runs the intake: every Canadian federal, provincial, and municipal tender feed, parsed at clause level, fit-scored against the firm's profile, and carried through a six-stage pursuit pipeline from Go/No-Go to Submit. Register runs the evaluation and drafting: it compiles the compliance matrix from the RFP, maps every clause against a provenanced clause library of past submissions, and drafts each volume with citations. Harbour decides what to pursue. Register decides whether the pursuit wins.

Primary procurement markets in scope are the federal CanadaBuys portal, the Quebec SEAO portal, the MERX mid-market vehicle, Ontario Tenders, and Alberta Purchasing Connection. The methodology ports to additional provincial portals, the MASH sector (municipal, academic, school-board, hospital), and pre-solicitation intelligence feeds; those extensions carry the same two engines with market-specific rule overlays.

Horizons run from the hour after a tender posts through the post-award debrief loop. Time granularity matches the decision at hand: hourly for opportunity intake and change-log tracking, daily for pursuit progression, release-to-submit for compliance matrix completeness, and weekly for win-loss pattern analysis.

02. The four surfaces Folio Bid reads

Every capture decision rests on four surfaces. Harbour reads the first two. Register reads the second two. The four together drive the bid from first look to signed submission.

The opportunity surface (Harbour). Every active Canadian tender across CanadaBuys, SEAO, MERX, Ontario Tenders, and Alberta Purchasing Connection, parsed into a single normalised index and refreshed on a two-hour cycle. Deduplicated by content hash. Every amendment, Q&A, extension, and cancellation surfaces as a change-log against the original solicitation.

The pursuit surface (Harbour). The six-stage pipeline every pursuit moves through: Assessment, Go/No-Go, Capture, Proposal Draft, Colour Team Review, Submit. Each stage owns its artefacts and sign-off trail so the post-award debrief has a complete record and the next pursuit starts with what the last one actually learned.

The clause surface (Register). Ten years of the firm's past submissions, structured as queryable prior art. Each clause answer carries provenance (which proposal, which author, which solicitation, won or lost, at what price) and a freshness score against current CanadaBuys standard clauses, the Trade Agreement implementation guide, and PSPC contract templates.

The compliance surface (Register). Every mandatory criterion, every point-rated criterion, every trigger clause parsed from the RFP and tracked Y or N or Partial with page references on both sides. The matrix is a live working artefact from Day 1, not a last-week deliverable. It is the artefact contracting authorities read first.

03. Harbour: opportunity intake and pursuit pipeline

Harbour is the part of Folio Bid that watches the market. It crawls the five primary Canadian procurement portals every two hours, parses each solicitation at the clause level, and scores every opportunity against the firm's NAICS codes, province footprint, contract value range, security clearance, and past-performance vehicles. Fit-scored opportunities land in a saved-search feed per capture lead. Nothing relevant is missed; nothing irrelevant reaches the inbox.

Set-aside detection is first-class. PSAB (Procurement Strategy for Aboriginal Business) triggers, OLA (Official Languages Act) French-service requirements, and CCP (Canadian Content Policy) thresholds are extracted from the solicitation and the evaluation grid automatically, with an evidence row per trigger. Bidders serving set-aside markets, Indigenous-owned firms, and bilingual captures stop missing obligations that used to disqualify them at evaluation.

Every pursued opportunity enters the six-stage pipeline: Assessment, Go/No-Go, Capture, Proposal Draft, Colour Team Review, Submit. Each stage owns structured artefacts, not freeform notes. The Go/No-Go stage carries the engine's verdict with confidence and factor list; the capture lead reads the reasoning and signs off. Colour Team reviews (Pink, Red, Gold) are captured in-system with named reviewers and comment trails. When the next pursuit needs the precedent, the artefact is retrievable in seconds rather than buried in email.

The post-award debrief loop closes the system. Win or loss outcome plus reason codes feed back into the clause library, the fit scorer, and the Go/No-Go factor weights. The next pursuit starts with what the firm actually learned, not with marketing boilerplate.

04. Register: compliance matrix, clause library, volume drafting

Register is the evaluation and assembly engine. It takes a live RFP and produces the three artefacts the contracting authority actually reads: a complete compliance matrix, a response-grade set of drafted volumes, and the cover and attestation pages that have to be exact.

The clause library is the foundation. Point the library at a folder of the firm's past RFx responses and it parses each document, extracts clause-level answers, and tags each by volume, agency, solicitation type, and outcome. Every entry carries provenance (source proposal, author, solicitation, won or lost, price) and a freshness score. The library maps Canadian equivalents of US Section L / M / C indices so mandatory criteria, point-rated criteria, and statement-of-work clauses are searchable as distinct categories. Winning firms already treat submissions as structured prior art that compounds across pursuits. Register makes that approach accessible to mid-market firms without a dedicated knowledge-management team.

The compliance matrix compiles itself as the engine parses the RFP. Every mandatory technical criterion (Go/No-Go) and every point-rated criterion is extracted from Annex A and the evaluation grid, numbered, and scored against the library. PSAB, OLA, and CCP triggers carry their own evidence rows. Each matrix row links to the RFP page where the clause lives and to the proposal page where the response sits. The evaluator does not have to search.

Volume drafting is AI-assisted, not AI-autonomous. The engine proposes clause-by-clause answers with citations back to the library. The capture lead reviews each answer, edits where the fit needs sharpening, and signs off. Matrix rows that were N become Y as the draft resolves. Under a compressed clock this approach is what keeps the bid submittable when narrative drafting alone cannot.

05. The Canadian procurement ruleset Folio Bid carries

Any credible capture tool has to know the Canadian procurement ruleset in the same way a grid-diligence tool has to know the NEM. Folio Bid's runtime carries an operating picture of every regime that decides bids before narrative is read.

Standard Acquisition Clauses and Conditions (SACC). Treasury Board Secretariat maintains the SACC manual; its general conditions, supplemental general conditions, and standard clauses appear in almost every federal solicitation. Register tracks the SACC version on each RFP and flags clause amendments that change exposure.

Canadian Content Policy (CCP). Federal solicitations above a threshold carry Canadian Content triggers that require a formal certification at bid close. The engine detects the trigger, flags the certification requirement, and pulls the evidence from the clause library where available.

Official Languages Act (OLA). Services-in-French requirements attach to specific solicitation locations, user populations, and service types. Harbour extracts the trigger and surfaces the French-service delivery obligations so bilingual capture teams (Quebec and national) can pull the right past-performance citations.

Procurement Strategy for Aboriginal Business (PSAB). Set-aside and Indigenous-benefit-plan triggers drive Register's compliance matrix where applicable. Indigenous-owned and PSAB-certified bidders receive the filter and evidence automatically.

Trade agreements. CFTA, CETA, CPTPP, and USMCA procurement chapters shift thresholds, notice obligations, and supplier-eligibility rules. The engine tracks which trade agreement governs a given solicitation and applies the right procedural checks.

06. Where bids actually die

Canadian mid-market firms lose bids for predictable reasons. Harbour and Register are engineered around those failure modes.

Mandatory review, not point-rated review. The primary disqualifier is an incomplete compliance matrix, not a weak narrative. A single unanswered mandatory criterion kills the bid before the evaluator reads page one of the technical volume. Register's live matrix closes this category.

Missed late breaks. Amendments, Q&A responses, and extensions land on the contracting authority's portal during the response window. A response built against the original release misses the clarification. Harbour's two-hour refresh and change-log surface every material edit against every pursued opportunity.

Canadian Content or OLA certification gaps. A federal solicitation above the threshold without the Canadian Content certification is disqualified at opening. An OLA-services solicitation without French-service staffing evidence fails the mandatory review. The Register trigger detection catches these before submission.

Compliance-matrix formatting. Contracting authorities disqualify matrices that are incomplete, unpaginated, or misaligned with the evaluation grid. The engine produces the matrix in the format the evaluator expects and links each row to both documents. The formatting problem stops being a problem.

Post-award debrief drift. Firms that do not run a debrief repeat the same losing pattern. Register's outcome-loop captures win or loss plus reason codes and refreshes the Go/No-Go factor weights so the next pursuit starts smarter.

07. Sample engagement shapes

Folio Bid is built to serve a handful of distinct engagement shapes. The patterns below are indicative rather than exhaustive: the engine runs Harbour and Register underneath each of them, with the deliverable and cadence adjusted to the shape.

Single-tender compliance matrix and bid pack. A capture lead has a named tender under a compressed clock and needs the compliance matrix compiled plus response volumes drafted to a capture-lead-signed submission. Target turnaround depends on RFP length and library coverage; a mid-market federal RFP is built to move from release to submittable pack inside the response window with review buffer. Output: signed compliance matrix, drafted technical and management volumes, pricing workbook stub, past-performance citations, trade agreement and set-aside certifications.

Ongoing capture subscription. A capture team running fifteen to eighty pursuits a year needs Harbour continuously watching CanadaBuys, SEAO, MERX, Ontario, and Alberta plus Register standing ready on every pursuit. Delivered as a recurring subscription with named capture-lead access, library refresh, Go/No-Go verdict surface, and compliance-matrix compilation on demand.

Clause-library build-out. A firm wants its last five to ten years of responses restructured into queryable prior art before running more pursuits through Register. Engagement shape: Register's ingestion pipeline points at the firm's archive, parses each response, tags clauses, establishes provenance, and publishes a freshness-scored library. One-off project feeding the ongoing capture subscription.

Indigenous-owned and PSAB-certified capture. An Indigenous business pursuing federal set-aside volume needs every PSAB, Indigenous-Benefit-Plan, and Indigenous-Content trigger surfaced automatically on the first pass with evidence rows populated from their own past performance. Harbour plus Register delivered with the PSAB-specific overlays enabled by default.

Quebec bilingual capture. A Quebec-based prime or national firm running bilingual SEAO pursuits needs French-language solicitation parsing, OLA trigger detection, and Quebec-specific contracting-vehicle indexing. Harbour and Register delivered in bilingual mode with Quebec evaluation-committee norms captured in the compliance-matrix format.

08. Response horizons

Different pursuits need different horizons. Folio Bid is built to run at the granularity the pursuit actually requires.

The hour after tender release. Harbour detects the new solicitation, parses it at clause level, scores its fit against the firm's profile, and surfaces the Go/No-Go input. Fit-scored opportunities are available to capture leadership within one refresh cycle of publication.

The working days inside the response window. Register compiles the compliance matrix on day one and updates it live as the draft evolves. Volume drafting iterates against the live matrix. Colour Team reviews capture their feedback in-system.

The submit window. Final clause-by-clause walk, attestation pages, cover letter, compliance-matrix export, pricing workbook export, and secure submission through the contracting authority's portal. Register produces the submission pack in the format each portal accepts.

The quarter after submission. The post-award debrief loop closes: win or loss outcome, reason codes, evaluation committee feedback where available, and library refresh based on what the engagement learned.

09. The market context Folio Bid carries

Every Canadian procurement engagement takes place inside a specific institutional context. Folio Bid's runtime carries the operating picture so the capture lead does not have to teach the engine the rules each time.

Treasury Board Secretariat sets the overarching procurement policy. Public Services and Procurement Canada operates most federal solicitations. Shared Services Canada runs IT-specific procurement. Provincial and territorial treasury boards operate the provincial portals with their own rulesets. The Office of Small and Medium Enterprises (OSME) supports bidders navigating the federal system.

The evaluation committee reads the compliance matrix first, the executive summary second, and the technical narrative third. This ordering is why Register prioritises the matrix above narrative drafting. Narrative quality matters; narrative without a complete matrix never gets read.

Rule changes land frequently. SACC version revisions, trade agreement threshold updates, PSAB policy amendments, OLA regulation updates. The engine has a named owner for tracking each regime and a change-log that feeds into the ruleset layer so response patterns stay current.

10. What AI-assisted capture actually changes

The direction of travel is well understood. Canadian government contracting is moving from spreadsheet-driven capture to AI-assisted capture that keeps a senior human in the seat. The change is not autonomy. The change is compression of the mechanical work that used to consume a capture team's entire response window so the senior humans can spend their time on the judgement calls that win bids.

Folio Bid's methodology accounts for this trajectory explicitly. AI-assisted clause retrieval and matrix compilation is one envelope. Human judgement on fit, win themes, pricing strategy, and capture narrative is a separate envelope. The engine surfaces where the boundary is rather than collapsing it. Submissions that try to cross the boundary (AI-autonomous drafting without capture-lead sign-off) lose bids because the narrative does not reflect the firm's actual differentiators. Submissions that keep the boundary clean compress response time without losing the judgement layer.

None of this changes the core question a capture lead is asking. It changes how much of the response window can go to answering it.