Skip to content
17_SPECIALIST_AGENTS · ONE_ACCOUNTABLE_HUMAN

One line in.
A defensible shortlist out.

Type the mandate. Mandate decomposes it, researches the company, drafts the role specification and builds the scoring model — then you approve the bar before a single candidate is scored against it.

Access is granted by approval, not by credit card. No trial, no self-serve tier.

AGENTS17
MODULES12
DIMENSIONS5
PERSPECTIVES4
INTELLIGENCE3-WAY
CALIBRATIONAUTO
01 / The problem

Executive search is judgment work performed under bad conditions.

Dozens of CVs. A hiring manager whose stated requirements drift from their revealed preferences. No consistent scoring. No audit trail. Decisions get made on recency and charisma, then rationalised afterwards.

From a slate you defend in the room —to a slate that defends itself in the minutes.
LIVE DEMO02 / Live

Type any role. Watch Mandate think.

Drop an executive brief — a one-liner, a paste from an email, a vague description. The Intake Agent decomposes it, calibrates a scoring model, and drafts the first three Boolean queries in real time.

SIMULATOR/api/demo

Structured mandate
Seniority
Director, reports to CIO
Function
IT operations, service management
Sector
Private healthcare, regulated
Scale signal
2,000–8,000 staff, multi-site
Context, web-grounded

Sector is consolidating; uptime and clinical-system availability dominate the operating agenda. Service-desk maturity and out-of-hours cover are the usual failure points at this scale.

14 sources consulted
Draft scoring model
Multi-site service operations26%
Clinical-system availability22%
Vendor & contract control18%

A draft. In the product you edit and approve it before anyone is scored against it.

ExampleType a real mandate above to run it live, or open an illustrative run:

NoteThis is what Mandate does in 30 seconds.

03 / Guardrails

What the system will never do.

These are enforced in the prompts, the application and the database — not stated as intentions.

  • Never decides for you

    No hire or no-hire verdict is produced anywhere in the product. Every output is decision support, with a human accountable for the call.

  • Never profiles a person

    No psychological labels, no protected-characteristic inference, no deception detection, and no audio, video, facial or voice analysis.

  • Never rewrites the record

    Approved artifacts cannot be edited — corrections create a new version and archive the old one. The audit log accepts inserts only.

04 / How it works

Calibration before evaluation.

The bar is set and approved before the faces appear. Everything downstream scores against the same model, and the model is versioned.

  1. Decompose

    One line becomes a structured mandate and a grounded company picture — the estate, the history, the constraints nobody put in the brief.

  2. Calibrate

    A structured intake produces weighted dimensions, must-haves and anti-patterns. You edit them. You approve them. Only then do they count.

  3. Evaluate

    Every CV is parsed, scored against the approved model and placed in a tier — with the evidence behind each dimension shown next to the number.

  4. Defend

    The shortlist carries its trade-offs. Client feedback recalibrates the model and re-ranks the field, and the change is versioned.

  1. Mandate
  2. Spec
  3. Calibration
  4. Sourcing
  5. Evaluation
  6. Ranking
  7. Shortlist
  8. Feedback
  9. Recalibrate
05 / Stack

The complete intelligence stack.

12 specialised modules across three intelligence layers. Every layer reads from the others — the system gets sharper the more you feed it.

ROLE_INTELLIGENCE

From one-line brief to ranked candidate slate.

  • Intake Agent
  • Job Spec Builder
  • Boolean Search
  • Sourcing Strategy
CANDIDATE_INTELLIGENCE

Read every candidate the way your best partner would.

  • CV Parsing
  • Evaluation Engine
  • Working-Style Signals
  • Positioning Kit
CLIENT_INTELLIGENCE

Predict what the client actually wants.

  • HM Portal
  • Culture Analysis
  • Feedback Interpretation
  • Weekly Reports
06 / The fusion layer

Three-way alignment, scored and evidenced.

Mandate cross-references candidate working-style signals, company culture, and hiring-manager preferences to produce a three-way alignment score — with anticipated objections and prepared responses.

Illustrative shape of the output — not live data

  • Candidate ↔ Company91
  • Candidate ↔ HM83
  • Overall alignment87
918387OverallCANDIDATESample CandidateCOMPANYSample CompanyHIRING MANAGERHiring Manager

Illustrative example — not live data

07 / Executive Intelligence · Add-on

Due diligence that survives the board minute.

A gated chain — success profile, interview plan, assessment — where each step requires human approval of the last. Approved records are immutable at the database layer. The audit trail is append-only. The assessment has no AI in it at all.

  • Success profileApproved · v3
  • Interview planApproved · v2
  • AssessmentHuman-authored
  • ReportUnlocks on approval
08 / Pricing

Pricing that scales with the practice.

All plans include the full agent stack, billed monthly. Billing begins once your workspace is approved — there is no self-serve signup.

  • Starter
    $399/mo

    1 user, 3 active searches

    • Full intelligence stack
    • Ranking, comparison and shortlists
    • 30-day evaluation history
    • Email support
    Request accessStarter
  • Agency
    $1,899/mo

    Unlimited users + searches

    • Everything in Growth
    • Global Executive Network
    • Custom skills + agents
    • Dedicated success partner
    • SLA + onboarding workshop
    Request accessAgency
  • Executive Intelligence
    Contact sales

    Add-on to any plan

    • Gated diligence chain
    • Versioned success profiles
    • Per-candidate interview plans
    • Immutable approved records
    • Append-only audit trail
    Contact salesExecutive Intelligence

    Scoped per engagement. Added to any plan above.

Full plan comparison, including what changes between tiers, is on the pricing page.

09 / Questions

Common questions.

Is my client data secure?

Every project is org-scoped. Row-level security in Supabase prevents cross-org access at the database layer; we don't rely on application-level filters alone. CV uploads sit in private storage with signed-URL access. We never train models on your data.

How is this different from LinkedIn Recruiter?

LinkedIn Recruiter is a sourcing surface — find people. Mandate is the operating system around it: intake, calibration, evaluation, feedback interpretation, shortlist generation, submission narratives. The Boolean queries we generate run inside LinkedIn; we don't replace it.

Does it source candidates for me?

No. It writes the LinkedIn, X-Ray and ATS queries — exact, broad, adjacent and competitor — and you run them. Candidates enter by CV upload. We would rather say that plainly than imply a funnel that does not exist.

Who is accountable for a decision?

You are. Generation never becomes operational truth — competency weights only take effect when a human approves the profile, and the approving identity is derived from the session, not supplied by the client.

Does it replace my judgment as a recruiter?

No, and the product is built to prevent it. Nothing produces a hire or no-hire verdict. Every output is decision support, every artifact is reviewed and approved by a person, and the approval is recorded with who did it and when.

Can I correct an approved record?

You create a new version. The previous one is archived and remains readable. Nothing is overwritten — the database rejects the edit regardless of role, including ours.

Why is there no free trial?

Access is granted by approval. Every workspace holds real candidate data under real obligations, and we would rather have the conversation first.

Does it work for in-house talent teams?

Yes. In-house teams use it for confidential and executive hires where the same rigour applies. The pricing is the same — a seat is a seat.
10 / Get started

Set the bar before you see the faces.

Tell us the mandate you are running now. A founder reads every request and replies within 48 hours, and we will walk a live search through with you before anything is billed.