One engine. Many roles. Content and configuration, not code.
Each worker is a definition, plus a knowledge scope, plus an asset scope, plus a set of adapters and grants. That is what allows a role as different as a collections specialist to run on the same runtime as an enterprise seller. Two flavours run on it today. The Inside Sales Assistant works enterprise sales development. The Property Assistant works sales as its primary service and rentals alongside it.
Inside sales and property share the kernel, the template, the tool schema, the scoring arithmetic and the compliance engine, byte for byte in the files that matter. What differs is the application layer, the ladders, the outcome contracts and the standing objectives. Vocabulary, briefings, escalation boundaries and disclosure phrasing per language are content, and content is where a flavour lives.
A domain with genuinely vulnerable counterparties required no engine change at all, only a different application layer, a different guardrail set and a different ladder. If that is the whole difference, the boundary is in the right place.
Each worker below is the same object: a definition the runtime executes. The difference between them is which ladder they pursue, which knowledge they draw on, and which systems they write back into.
Navigates organisations, follows referral chains, detects initiatives before they become formal procurements, qualifies progressively against your methodology, and books meetings with the people who actually decide. Eight ladders, one outcome contract, eleven standing objectives.
Explore inside salesarrow_forwardAnswers enquiries, qualifies buyers and applicants, arranges accompanied viewings, wins market appraisals, reports back to owners, and establishes intention on tenancies ending. Fifteen ladders across five outcome contracts, and a fair treatment review that gates go-live.
How portability worksarrow_forwardDeal guidance, call preparation, post call debrief, longitudinal skill analysis, and objection rehearsal against a simulated contact. Grounded in governed methodology rather than generic advice.
How coaching worksarrow_forwardTracks contract expirations well ahead of the date, monitors utilisation for both churn risk and expansion signal, notifies stakeholders when releases ship, and steps in when support escalates.
Systems it connects toarrow_forwardTakes an asset request raised by another worker, drafts the one pager or localised deck from governed knowledge, and submits it for approval before publication. The point at which a hub becomes a fleet.
The request boardarrow_forwardReview meetings, adviser follow up and information packs, over licensed wealth management domain content and a core banking system of record.
Domain contentarrow_forwardArrangement agreed, hardship review, specialist referral. A ladder where the ethical guardrails matter more than the conversion, and where the platform's consent machinery earns its place.
Guardrailsarrow_forwardTerms confirmed, clarification meetings, escalation, against an ERP and a contract repository rather than a CRM.
Adapter modelarrow_forwardDrift detected, cluster triaged, revision proposed. A supervisory worker whose authority is a strict, enumerated subset, and which may never approve its own proposals.
Supervision modelarrow_forwardRun the inside sales worker and you are in a sales application. Run the property worker and you are in a property system. Same engine, same screens, different product.
That is not a theme. The worker pack carries an application profile: the name of the application, the vocabulary, the navigation for each role, and the iconography. The runtime renders from that profile rather than from labels hardcoded by us. Account becomes Property. Contact becomes Buyer, Vendor, Applicant, Landlord or Tenant, five counterparties rather than one. Meeting becomes Viewing.
A lexicon that carries a lifecycle does not flatten it. Where the person changes name as they move through the funnel, the lexicon says so. Mapping everyone to their end-state term has the worker greeting an enquirer as though the relationship already existed.
Neither picture is a translation of the other. The ladders end somewhere different: a meeting with the CIO in one, a viewing confirmed or a market appraisal booked in the other. The outcome contracts capture different fields, and property declares five of them where inside sales declares one. Even the worst failure the calibration loop guards against is different: offering a callback into resistance in one, and in the other, giving a figure on the phone.
The property worker does not value and it does not price. Not a figure, not a range, not a guide, not a number qualified into safety. A figure given on the phone becomes the number every later conversation is measured against, so refusing turns into the reason to book the appointment, which was the outcome the call wanted. The appraisal appointment is the outcome, and it scores full marks.
A tenant may extend the vocabulary but not replace it. Adding a term the pack does not define is configuration. Redefining one it does define breaks every measure bound to it and every future upgrade, so the editor refuses rather than warns.
Your own brand sits on top and is preserved across upgrades: your name, your mark, your accent colour, your outbound identity. The profile belongs to the pack; the branding belongs to you.
| Worker | Primary ladder | Systems of record | Domain content |
|---|---|---|---|
| Inside Sales Assistant | Executive meeting, then stakeholder meeting, then account team follow up | CRM, calendar, workspace | Sales methodology, product and competitive knowledge |
| Property Assistant | Five contracts. Viewing confirmed for a buyer or an applicant, market appraisal booked for a vendor, landlord decision confirmed, renewal agreed for a tenant | Instruction record, listing platform, enquiry and viewing workflow | Local market playbook, instruction and transaction knowledge |
| Domain Coach | Skill gap identified, then rehearsal completed, then measurable improvement | The Hub's own evidence base, plus CRM | Methodology library, objection handling, conversation examples |
| Customer Success | Renewal secured, then conversation scheduled, then escalation routed | CRM, entitlement and usage systems, service desk | Product, entitlement and support knowledge |
| Marketing | Asset published, then draft in review, then brief clarified | Asset library, brand system | Brand, positioning and content knowledge |
| Relationship Assistant | Review meeting, then adviser follow up, then information pack | Core banking, portfolio and CRM systems | Licensed wealth management content |
| Collections & Workout | Arrangement agreed, then hardship review, then specialist referral | Servicing and collections platform | Regulatory and hardship handling content |
| Fleet Manager | Drift detected, then cluster triaged, then revision proposed | The Hub itself | Scoring rubrics and calibration policy |
The platform adapts to your existing infrastructure rather than requiring you to change your workflows. Knowledge and intelligence population run in parallel over a longer period, and deepen continuously after launch.
Onboarding is a sequence of wizards, not a form. Twelve for inside sales, thirteen for property, which adds a fair treatment review. Completion is a state rather than an event: any wizard can be left and resumed, and each says what it blocks, so it is always clear which outstanding item is holding up a date.
Voice first, real time, multilingual, regulated, and commercially measurable. A runtime that carries enterprise inside sales and property, with their different counterparties and their different duties, carries almost anything.