Interview training
that measures
as it teaches
Elicit places your investigators and enforcement officers in conversation with a realistic AI subject. Every question they ask is logged, typed, and scored against the PEACE framework. Every session ends with an assessment report that tells you exactly where each officer stands - and what they need to work on next.
You cannot improve what you cannot measure
Every police force trains to PEACE. Most invest significantly in roleplay courses, professional actors, and facilitated sessions. Very few can tell you, officer by officer, how well each one applies the framework under pressure - because current training produces no objective record.
"The quality of investigative interviewing is one of the most significant variables in the criminal justice process. It determines what evidence is captured, how reliable it is, and whether a prosecution can succeed."
College of Policing, Authorised Professional Practice: Investigative InterviewingRole-play does not scale
A professional actor handles six to eight sessions per day. Your force has hundreds of investigators at different PIP levels. The scheduling alone makes meaningful regular practice impossible for most officers.
eLearning is passive
Video-based modules tell officers what good interviewing looks like. They do not develop the skill of doing it. Watching a model interview and conducting one under pressure are entirely different competencies.
No objective assessment exists
Facilitator feedback is subjective, inconsistent, and unrecorded. There is no standard by which a detective sergeant can compare officer A and officer B. Development planning is based on impression, not evidence.
HMICFRS scrutiny is increasing
Inspection frameworks now examine whether forces have systematic approaches to developing and assessing investigative skills - not just whether training was delivered. The absence of a documented, objective assessment process is increasingly a finding in itself.
Industry rates for professional roleplay actors and training facilitation. Officer time costed at detective constable top-of-scale rate. Elicit sessions are available on demand, with no scheduling, no actor fees, and a full assessment report included in each session.
A flight simulator for investigative interviewing
Elicit is an AI-driven interview training platform for investigators, enforcement officers, and anyone whose job requires eliciting accurate accounts from people who may not want to give them. It uses realistic AI subjects with structured knowledge graphs - facts they know, facts they will only reveal under the right conditions, and facts they will not give up without evidence challenge.
A pilot does not learn to land in fog on the first flight. They practice in a simulator until the response is automatic. Elicit applies the same logic to interviewing. The AI subject is as realistic as your officers can handle. The assessment is more detailed than any human observer could provide. And the subject is available at any time, never has a bad day, and never gives the same performance twice.
- Interviews conducted by voice, with a realistic AI avatar - not multiple-choice or video
- Every session assessed in real time against the PEACE framework
- Knowledge graph tracks exactly which facts were elicited, how, and when
- Question types classified automatically - open, closed, leading, multiple, forced-choice
- Rapport score measured across the session using the ORBIT framework
- Every session produces a full assessment report for the officer and their supervisor
Officer receives the scenario briefing
Before the interview begins, the officer reads a scenario briefing - the nature of the incident, any background information they would realistically hold, and the objectives for the interview. No more, no less than they would have in a real case.
The interview begins
The AI subject - a realistic video avatar with a voice and a fully scripted persona - enters the room. The officer conducts the interview by voice. The AI responds in character, tracks the officer's approach, and reveals information according to its knowledge graph only when the right technique is applied.
Live feedback (optional)
If enabled by the supervisor, a live feedback panel shows the officer's PEACE phase, question type in real time, and rapport markers as they appear. In assessment mode this panel is hidden - the officer works without the guide.
Assessment report generated
On completion, the Elicit Assessment Report is produced immediately. PEACE phase breakdown by time, question type analysis, information yield score, rapport score, and a cited narrative assessment with specific examples from the session transcript.
A growing scenario library. Four flagship subjects.
Every scenario comes with a structured knowledge graph - facts the subject knows, facts they will only reveal under the right conditions, and facts they will not give up without an evidence challenge. The interview score is a direct measure of how much of that knowledge the officer managed to elicit, and how they did it.
The Cooperative Witness
A young person who witnessed a street robbery. Nervous, easily led, prone to speculation. The scenario tests whether the officer can build rapport quickly, use open questions to elicit a structured account, and avoid contaminating the witness's memory with suggestion or leading language. Deceptively difficult for officers new to PEACE discipline.
The Reluctant Witness
A bystander who saw more than they are letting on. Not hostile, but afraid of involvement. The AI subject shuts down in response to pressure and responds only to patience and a non-confrontational approach. The scenario tests engagement technique, silence management, and the officer's ability to sustain rapport without rewarding evasion.
The Evasive Suspect
Articulate, well-prepared, experienced with police interviews. The AI suspect deflects, qualifies, and introduces counter-narratives. Evidence must be introduced at the right moment - too early and it is absorbed; too late and the officer has lost the thread. The scenario tests structured evidence challenge, timeline probing, and no-comment management.
The Hostile Suspect
Verbally combative, personally offensive, testing the officer's professional composure at every exchange. The AI subject escalates in direct response to any loss of control from the officer. The scenario tests de-escalation under sustained pressure, maintaining the interview structure when the subject is trying to derail it, and professional closure regardless of outcome.
The Elicit Assessment Report
Every completed session generates an Elicit Assessment Report - a structured, objective record of the officer's performance across the full PEACE framework. Not a pass or fail. A forensic breakdown of what happened, when, and how well.
The report is written for two audiences simultaneously: the officer, who needs specific, actionable feedback; and their supervisor, who needs a comparable, objective basis for development planning and PIP progression evidence.
The investigating officer
Receives specific, timestamped feedback on each PEACE phase, with examples drawn from the transcript. Knows exactly which questions worked, which contaminated the account, and where the information yield was lost.
The detective sergeant or supervisor
Receives comparable, objective data across their whole team. Can identify the officers who are ready to progress and the ones who need structured support - without relying on impression or anecdote.
The L&D and force training team
Gets aggregate performance data across scenarios and PIP levels. Can evidence training activity, track improvement over time, and demonstrate a systematic approach to interview quality development for HMICFRS and the College of Policing.
Assessed against the standards your force already uses
Every Elicit scenario, question classification, and report dimension is mapped to frameworks the College of Policing, HMICFRS, and your force's own training curriculum already recognise. There is no proprietary scoring system to explain or justify.
The College of Policing's PEACE model is the mandated framework for investigative interviewing across all 43 territorial forces. Elicit maps every session to each phase, with time-in-phase tracking and phase-specific scoring.
- Preparation and planning
- Engage and explain
- Account - free recall, clarify, challenge
- Closure
- Evaluation
The PIP framework sets competency standards for investigators at Tier 1 (volume crime), Tier 2 (serious and complex), and Tier 3 (major and critical). Elicit scenarios and reports are aligned to PIP tier requirements.
- Tier 1 - volume crime investigators (Cooperative Witness)
- Tier 2 - serious crime investigators (Evasive and Reluctant scenarios)
- Tier 3 - major crime investigators (Hostile Suspect)
- Evidence suitable for PIP portfolio and progression documentation
Rapport is one of the most significant variables in interview outcome. Elicit measures rapport using the ORBIT framework across five dimensions, producing a rapport trajectory across the session and a score cited to specific exchanges.
- Overview - orientation and scene-setting
- Rapport - quality of relationship established
- Basic rules - framework compliance and explanation
- Introduction - transition to the account
- Topic handling - responsiveness and follow-through
The questions you will need to answer
When HMICFRS asks about interview quality, or a detective superintendent asks why two officers at the same PIP level perform so differently, Elicit gives you something concrete to show.
"How do you assess the quality of investigative interviewing across your force?"
The Elicit Assessment Report constitutes an objective, comparable, session-by-session record of each officer's PEACE framework application. No other training method produces this record. Aggregate data across your force gives inspectors the systematic approach they are looking for.
"How do I know which of my team needs development before the next major inquiry?"
The supervisor dashboard shows every officer's session history, PEACE phase scores, and information yield across scenarios. You can identify the officers who are ready and the ones who need targeted support - based on data, not impression.
"What is the cost model and how does it compare to our current programme?"
Elicit sessions are available on demand, with no actor fees, no scheduling overhead, and no venue costs. At scale, Elicit replaces the most expensive element of your current provision while producing a level of assessment data that role-play cannot match. Contact us for force-level pricing.
See Elicit demonstrated
We run live demonstrations for training leads, detective superintendents, and force L&D teams. The demonstration covers a complete officer session from briefing to assessment report, with a discussion of how Elicit fits your current training architecture and PIP framework.