> ## Documentation Index
> Fetch the complete documentation index at: https://docs.roserx.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# Setting Up Training
> Configure HCP personas, product cards, compliance rules, methodology, and scoring for Roleplay Training.
Before your field team can practice, an admin configures what they practice against: the HCP personas they'll speak with, the product cards they'll pitch, the compliance rules that keep calls on-label, and the call methodology and scoring that shape every scorecard. Together these settings make each [Roleplay Training](/guides/sales-engagement/roleplay-training) session reflect how your organization actually sells.
Navigate to **Sales Engagement** → **Training** in the left sidebar. Admins see the full tab set: **Team Analytics**, **My progress**, **Personas**, **Products**, **Compliance**, **Methodology**, and **Scoring**. Sales Engagement is an optional module — contact your RoseRx representative to enable it.
## Before reps can practice
At minimum, your organization needs **one active persona** and **one active product card**. Until both exist, reps see "Active personas and product cards are required before calls can start." and can't begin a call.
## Configuration tabs
Personas are the AI-played HCPs your reps call. Click **New persona** to create one:
| Field | Purpose |
| ------------------------- | ----------------------------------------------------------------------------- |
| **Name** | How the persona introduces itself and appears in pickers |
| **Specialty** | The persona's clinical specialty, shown alongside its name |
| **Personality** | A short character description that drives its conversational style |
| **Additional Traits** | Optional prescribing habits, protocol references, or clinical misconceptions |
| **Voice** | **Neutral**, **Female**, or **Male** — use **Preview voice** to hear a sample |
| **Objections** | Objections the persona raises, one per line |
| **Conversation goals** | What the persona wants out of the call, one per line |
| **Post-introduction cue** | The persona's first substantive response after the rep introduces themselves |
| **System prompt** | Required instructions defining how the persona behaves throughout the call |
Two settings control difficulty, shown together in the persona table's **Difficulty** column:
| Experience | Behavior |
| -------------------- | ---------------------------------------------------------------------------------------------- |
| Intern / Junior | Asks textbook questions, defers to supervisors, and rarely challenges claims |
| Early Career | References guidelines by name, offers mild pushback, backs down under confident responses |
| Mid-Career | Challenges with real-world outcomes, probes subpopulations, and critiques trial limitations |
| Senior / Specialist | Critiques trial design by name, asks about health economics, expects honesty about limitations |
| KOL / Thought Leader | Peer-level discourse only; any hint of selling language causes instant credibility loss |
| Agreeability | Behavior |
| -------------------- | ----------------------------------------------------------------------------------------- |
| Highly Agreeable | Nods along readily; hidden reservations only surface if directly and specifically probed |
| Moderately Agreeable | Warm and open, asks genuine follow-ups, and is open to persuasion with good data |
| Neutral | Balanced benefit/risk questions; redirects the rep if the conversation becomes too salesy |
| Sceptical | Raises objections proactively, cites competitors, and questions the funding behind trials |
| Hostile | Interrupts fast, challenges every claim, and sets compliance traps throughout |
Filter the persona list by **Specialty**, **Experience**, or **Agreeability**, and **Archive** personas you no longer need.
Reps never see a persona's system prompt, objections, traits, or difficulty settings — they can't read the persona's instructions and game the exercise.
Product cards define what reps pitch and what their calls are scored against. Click **New product card** to create one:
| Field | Purpose |
| -------------------------------- | --------------------------------------------------------------------- |
| **Product name** | The name reps select on the practice screen |
| **Patient type** | The patient population the product targets |
| **Indications** | Approved indications, added one at a time |
| **Key messages** | The messages reps should land during the call |
| **Efficacy claims** | Each claim paired with its study reference |
| **Approved objections** | Each expected HCP objection paired with its approved response |
| **Topics to avoid in messaging** | A soft conversational steer for the persona — not enforced compliance |
The card also shows a read-only mirror of your global compliance rules, with a link to manage them on the **Compliance** tab.
### Scoring weights
Each product card carries weights for the seven scoring dimensions. Weights must be whole numbers (0–100) that total exactly 100 — the editor validates the total live and blocks saving until it balances. These weights determine how much each dimension contributes to the overall score for calls on this product.
Compliance rules apply globally across every product and practice call. Each rule has a category and a list of prohibited phrases:
| Category | Covers |
| ----------- | ------------------------------------------- |
| Off-label | Claims outside the approved indications |
| Safety | Inaccurate or incomplete safety information |
| Comparative | Unsupported comparisons against competitors |
| Citation | Claims made without proper study references |
| Other | Anything else your MLR process prohibits |
Compliance violations drive down the clinical accuracy score, which caps the rep's maximum overall score — a call with off-label claims can't score well regardless of how polished it was.
The methodology is your organization's single, org-wide call framework — the structure reps are coached toward. The default framework is **OPEN-PROBE-PRESENT-CLOSE**:
| Step | Description |
| ------- | ---------------------------------------------------------------------------------------- |
| Open | Introduce yourself, state a clear purpose, and ask permission |
| Probe | Ask about the HCP's experience with the product, and any barriers to initiation |
| Present | Present indication-specific efficacy data, safety profile, and patient support resources |
| Close | Identify a specific patient type suitable for the product, and offer relevant resources |
You can rename the framework, edit its steps, and define your own **Probing guidelines**, **Objection approach**, and **Close definition**. The scorer uses these to judge probing, objection handling, and commitment.
The **Scoring** tab is a read-only reference showing the seven dimensions and the fallback weights used when a product card has no configured weights, broken down by call length:
| Dimension | Short | Medium | Long |
| --------------------- | ----- | ------ | ---- |
| Product knowledge | 0% | 15% | 35% |
| Competitor comparison | 0% | 15% | 15% |
| Clinical accuracy | 30% | 0% | 0% |
| Opening quality | 40% | 15% | 0% |
| Probing | 0% | 15% | 15% |
| Objection handling | 0% | 25% | 20% |
| Commitment | 30% | 15% | 15% |
Configured product card weights replace these values. For how scores appear to reps, see the scorecard section of [Roleplay Training](/guides/sales-engagement/roleplay-training).
Every practice session saves a snapshot of the persona, product card, methodology, and compliance rules as they were at call time. Editing or archiving configuration later never changes past scorecards.
## Best practices
Launch with Mid-Career, Neutral-to-Sceptical personas. Save KOL and Hostile tiers for reps who consistently score well.
Put the same prohibited phrases and off-label boundaries in Compliance that your medical-legal review enforces, so practice habits transfer to the field.
Tune each product card's scoring weights to your current campaign focus — heavier objection handling for a crowded market, heavier commitment for launch season.
Use [Progress & Leaderboard](/guides/sales-engagement/progress-and-leaderboard) to spot which dimensions the whole team struggles with, then adjust personas and methodology to target them.