How to train the lettings team to review AI agent decisions

Lettings negotiators should be trained to review what the agent does before a tenancy starts: applicant screening, referencing summaries, offers to landlords and tenancy paperwork. Teach how it scores and holds, calibrate on real applications, shadow live lettings with every acceptance checked, and sign off negotiators who never let a legal check or fairness question pass.

What does the lettings team review?

A lettings agent typically answers enquiries, books viewings at depth 3, collects application details, gathers references and affordability evidence, drafts a referencing summary and a recommendation for the landlord, prepares tenancy agreements from templates, and schedules move-in tasks. It holds anything outside its bounds: incomplete referencing, affordability below the landlord's threshold, guarantor arrangements, special clauses, and every final acceptance.

Lettings brings two risks the team must be trained for. Fairness: decisions about who gets a home must not discriminate, directly or through proxies such as postcode or name. And legal checks: many markets require specific checks before a tenancy starts, and the details vary by country. The agent can gather evidence for them; the negotiator confirms they are complete.

The lettings manager owns the programme. The compliance or legal adviser for the agency agrees which checks are mandatory in each market where it operates.

What does the programme look like, step by step?

1. Learning, week 1: negotiators complete lessons on how the agent screens and summarises, its bounds, and which outputs are drafts versus actions.
2. Agent, depth 1: builds a practice set of past applications with referencing summaries and outcomes, anonymised.
3. Calibration, week 2: negotiators decide alone whether to recommend each applicant, and why. Compare. Look closely at any case where recommendations split on something other than the evidence.
4. Shadow, weeks 3 and 4: negotiators handle live applications; a senior negotiator checks every recommendation and every tenancy pack before it goes out.
5. Agent, depth 2 and 3 in live work: books viewings, drafts summaries and paperwork, holds acceptances.
6. Sign-off: lettings manager assesses shadow decisions on accuracy, fairness and completeness of legal checks.
7. Ongoing: monthly calibration, and a refresher whenever templates or legal requirements change.

What must negotiators learn to check?

Affordability evidence. Payslips and bank statements uploaded by the applicant are external until verified. A reference from an employer reached through a number found independently is stronger. Negotiators check what the summary rested on.

Referencing gaps. A missing previous landlord reference, a short employment history, a guarantor whose own evidence is thin. The agent flags; negotiators decide what more to ask.

Fairness. Would the recommendation have been the same for an applicant with a different name, nationality, family situation or source of income, all else equal? Calibration should include paired cases to test this. Recommendations must rest on the landlord's lawful criteria and the evidence.

Mandatory checks. Whatever the law in your market requires before letting, such as identity or eligibility checks, deposit handling or safety documentation. Negotiators confirm each is done and recorded.

Tenancy paperwork. Correct parties, rent, dates, deposit terms and any special clauses agreed with the landlord. Templates are bounds; special clauses are where errors creep in.

Landlord instructions. A landlord's preference is not automatically lawful. Negotiators learn to recognise instructions they must not follow and escalate them.

Paired cases make fairness concrete. Present two applications for the same flat with the same income, references and deposit; one applicant receives part of their income from benefits, the other entirely from employment. If negotiators recommend them differently, the group discusses why, and whether the reason rests on the landlord's lawful criteria or on assumption. Where the agent's draft treated them differently, the agent's owner is told the same day.

Which controls leave evidence?

• Lessons before shadow. Evidence: completion dates.
• Every recommendation reasoned. Evidence: reasons tied to evidence and the landlord's criteria.
• Mandatory checks recorded. Evidence: a completed check list on each let.
• Tenancy packs checked before issue. Evidence: reviewer's name.
• Unlawful instructions escalated. Evidence: escalation log.

What should you measure?

• Shadow accuracy per negotiator.
• Referencing re-dos after a recommendation.
• Paired-case consistency in calibration.
• Missing mandatory checks found in audit.
• Paperwork errors found after signing.

Where does it usually go wrong?

Speed beats checks. In a fast market, negotiators push applications through to secure the let. Mandatory checks must be a hard stop.

Fairness is assumed. Without paired-case calibration, bias stays invisible.

Special clauses slip. The template is right; the added clause is wrong. Read them every time.

Landlord pressure goes unrecorded. Log it, with the date, the instruction and what the negotiator did next.

Practise what happens when something goes wrong with an AI agent incident drill for the lettings team. New negotiators use the reviewer onboarding checklist for the lettings team, and skills are set out in the AI agent competency matrix for the lettings team. Practice teams run the same stages; see how to train the practice team to review AI agent decisions. The full programme is in training teams for AI agents.

Where the Academy fits

Pryme Intelligence Academy's Operator track teaches negotiators how a governed agent decides and how to review its drafts, with module tests set in the learner's own industry and role, so they practise on lettings cases. Lessons can be listened to as audio, which suits negotiators who spend much of the day between viewings.