> ## Documentation Index > Fetch the complete documentation index at: https://learn.nexudus.com/llms.txt > Use this file to discover all available pages before exploring further. # Agent Types Reference > Comprehensive reference guide for all available proactive AI agents, what they monitor, when they trigger, and recommended configuration settings. ## Overview This page provides detailed information about every available proactive agent in Nexudus. Each agent monitors different patterns in your coworking space and proposes actions tailored to that situation. Agents are organized by category below. Some agents are **Live** and available now; others are marked as **Coming Soon** and will be available in future releases. Use this page to: * Understand what each agent watches for * Decide which agents are right for your space * Learn recommended configuration settings * Understand what variables are available for customization * See which agents are safe for auto-execute Enable and customize each agent's behavior. Step-by-step guide to agent configuration. Review and approve proposed actions. *** ## Quick Reference Table | Agent | Category | Status | | ------------------------------------------------------------ | ---------- | --------- | | [Due Invoice Reminders](#-due-invoice-reminders) | Billing | ✅ Live | | [Contract Renewal](#-contract-renewal) | Billing | ✅ Live | | [Support Pattern Detection](#-support-pattern-detection) | Support | ✅ Live | | [Support Category Spike](#-support-category-spike) | Support | ✅ Live | | [FAQ Gap Detection](#-faq-gap-detection) | Knowledge | ✅ Live | | [Lead Follow-up](#-lead-follow-up) | Sales | 🔜 Coming | | [Onboarding Nudges](-#onboarding-nudges) | Engagement | 🔜 Coming | | [Tour Reminders](#-tour-reminders) | Sales | 🔜 Coming | | [Tour No-Show Follow-up](#-tour-no-show-follow-up) | Sales | 🔜 Coming | | [Booking No-Show Follow-up](#-booking-no-show-follow-up) | Engagement | 🔜 Coming | | [At-Risk Member Detection](#-at-risk-member-detection) | Retention | 🔜 Coming | | [Upsell Opportunities](#-upsell-opportunities) | Revenue | 🔜 Coming | | [Cancellation Risk Detection](#-cancellation-risk-detection) | Retention | 🔜 Coming | | [Lead Qualification Scoring](#-lead-qualification-scoring) | Sales | 🔜 Coming | *** ## Billing Agents ### 🟢 Due Invoice Reminders **Status:** ✅ Live\ **Category:** Billing\ **Channel:** Email (default), WhatsApp, Voice supported\ **Auto-Execute Safe?** ✅ Yes #### What It Does Monitors unpaid invoices and sends friendly payment reminders when invoices become overdue. Escalates through multiple reminder tiers based on how long the invoice has been outstanding. #### When It Triggers * **Tier 1 (Friendly Reminder):** Invoice is 3+ days overdue (default) * **Tier 2 (Follow-up):** Invoice is 7+ days overdue (default) * **Tier 3 (Critical Escalation):** Invoice is 14+ days overdue (default) * Tiers are fully configurable via **Days Overdue — Tier 1**, **Tier 2**, and **Tier 3** settings * Tier 3 triggers regardless of minimum amount; Tiers 1–2 respect **Minimum Due Amount** threshold * Only non-draft, non-void invoices with a past due date #### What It Monitors * Invoice status and due dates * Payment history patterns * Outstanding balance amounts * Previous reminder history #### Message Content Each reminder includes: * Friendly greeting with customer name * Outstanding invoice details (amount, due date, invoice number) * Payment link for immediate settlement * Contact information if the customer has questions * Escalating urgency based on tier #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------- | ------------------------------------------------------ | | **Cooldown** | 24–48 hours | Prevents duplicate reminders for same invoice | | **Max Actions/Day** | 20–50 | Scales with your customer base size | | **Auto-Execute** | ✅ Enabled | Safe for all tiers after testing period | | **Confidence Threshold** | 70–85% | Auto-generated per candidate (0.70–0.95 based on tier) | #### Configuration Parameters | Parameter | Default | Description | | ----------------------------------------------- | ------- | ----------------------------------------------------------------------------- | | **Minimum Due Amount** | `0` | Invoices below this amount are ignored for Tiers 1–2. Tier 3 always triggers. | | **Days Overdue — Tier 1 (Friendly Reminder)** | `3` | Days past due before a friendly payment reminder | | **Days Overdue — Tier 2 (Follow-up)** | `7` | Days past due before a follow-up with urgency | | **Days Overdue — Tier 3 (Critical Escalation)** | `14` | Days past due before a critical escalation | #### Available Variables Use these in custom prompts or voice messages: * `{{customer_name}}` — Full name of the customer * `{{customer_email}}` — Customer email address * `{{invoice_number}}` — Invoice reference identifier * `{{invoice_amount}}` — Outstanding balance formatted with currency (e.g. `USD 150.00`) * `{{invoice_description}}` — Invoice description * `{{due_date}}` — Original due date (formatted as `MMMM dd, yyyy`, e.g. `June 14, 2026`) * `{{days_overdue}}` — Number of days past due * `{{reminder_tier}}` — Current escalation tier (1, 2, or 3) * `{{company_name}}` — Name of your space/location * `{{session_id}}` — Unique session identifier for voice calls #### Best Practices * Enable immediately for all locations with invoicing * Start with manual review for the first 10 actions to verify quality * Monitor payment response rates weekly * Adjust cooldown if customers report too many reminders * Consider increasing confidence threshold to 90% for auto-execute once comfortable * Set **Minimum Due Amount** to filter out small invoices from automated reminders (Tier 3 still triggers for critical cases) * Proposed actions expire after 3 days if not reviewed — consider reviewing older candidates promptly *** ### 🟢 Contract Renewal **Status:** ✅ Live\ **Category:** Billing\ **Channel:** Email (default), WhatsApp, Voice supported\ **Auto-Execute Safe?** ✅ Yes #### What It Does Proactively reaches out to customers when their contract is approaching expiry, with context about their tenure, plan details, and pricing to start a renewal conversation. Helps prevent churn by engaging members before their contract ends. #### When It Triggers * When contract expiry falls within the configured number of days (default: 30 days) * Only for active contracts with a price * Contracts that are not cancelled and have a valid contract term * Not sent if a recent action has already been proposed for the same contract or member #### What It Monitors * Contract end dates (ContractTerm field) * Customer tenure and start date * Contract pricing and plan details * Contract active and cancellation status * Previous outreach cooldown per member #### Configuration Parameters | Parameter | Default | Description | | ---------------------- | ------- | -------------------------------------------------------------------------------------------------------- | | **Days Before Expiry** | `30` | How many days before contract expiration the agent starts proposing renewal outreach. Range: 1–180 days. | #### Message Content Each renewal outreach includes: * Personal greeting acknowledging the member's time with you * Contract end date reminder with formatted date * Summary of their tenure (total days as a member) * Plan name and pricing details * Invitation to discuss renewal needs * Non-pushy, value-focused messaging (avoid mentioning price or discounts in custom prompts) #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------------- | ----------------------------------------------------------------- | | **Cooldown** | 72–168 hours (3–7 days) | Allows multiple touches over renewal window | | **Max Actions/Day** | 10–15 | Realistic for renewal conversations | | **Auto-Execute** | ✅ Enabled | Safe for relationship-building after testing | | **Confidence Threshold** | 80–97% | Increases as expiry approaches (0.85 base, up to 0.97 for urgent) | #### Available Variables Use these in custom prompts or voice messages: * `{{customer_name}}` — Full name of the customer * `{{customer_email}}` — Customer email address * `{{contract_name}}` — Current plan or membership name * `{{expiry_date}}` — When contract expires (formatted as `MMMM dd, yyyy`, e.g. `July 15, 2026`) * `{{days_until_expiry}}` — Number of days remaining until contract expires * `{{start_date}}` — When they joined your space (formatted as `MMMM dd, yyyy`) * `{{tenure_days}}` — Total days as a member * `{{contract_price}}` — Contract price formatted with currency (e.g. `USD 299.00`) * `{{contract_quantity}}` — Number of contracts/units * `{{company_name}}` — Name of your space/location (for voice messages) * `{{session_id}}` — Unique session identifier (for voice messages) #### Prompt Guidance When customizing prompts, focus on: * The value the member has received during their tenure * A helpful, non-pushy tone about renewal * **Do not** mention price or discounts — the operator handles negotiation #### Best Practices * Enable for all locations with active contracts * Test first 5–10 renewals manually before enabling auto-execute * Personalize the first message for VIP or high-value customers * Monitor renewal conversion rates as a key performance indicator * Consider adding promotional incentives in custom messages during slower periods * Priority increases as expiry approaches: Critical (≤7 days), High (≤14 days), Medium (≤21 days), Low (>21 days) *** ## Support Agents ### 🟢 Support Pattern Detection **Status:** ✅ Live\ **Category:** Support\ **Channel:** Email (default), WhatsApp, Voice supported\ **Auto-Execute Safe?** ❌ No #### What It Does Detects when a customer has submitted an unusually high number of support tickets within a configurable timeframe and proposes a proactive check-in to address underlying issues. The agent evaluates ticket frequency, priority levels, and open status to determine the appropriate outreach confidence and urgency. #### When It Triggers * Customer submits N+ support tickets within a configurable lookback window (default: 3 tickets in 14 days) * Tickets are not already handled by AI chat sessions * No recent proactive outreach has occurred for this member (deduplication by week) * Daily action rate limit has not been reached #### What It Monitors * Support ticket frequency and recency within the lookback window * Ticket priority levels (High and Critical increase confidence) * Open vs. closed ticket status (open tickets indicate unresolved issues) * Ticket subjects (top 3 most recent, used in criteria payload) * Member's historical outreach cooldown status #### Configuration Parameters | Parameter | Default | Description | Range | | ------------------------ | ------- | ------------------------------------------------------------------ | ------------ | | **Minimum Ticket Count** | `3` | Number of support tickets a member must submit to trigger outreach | 1–50 tickets | | **Lookback Window** | `14` | Number of days to look back for ticket activity | 1–90 days | #### Message Content The proposed outreach includes: * Empathetic acknowledgment of recent support interactions * Offer to help resolve any ongoing concerns * Assurance that their experience matters * Focus on listening, not listing specific ticket details or technical information **Prompt Guidance:** The system prompt emphasizes empathy and offering help. Do not list specific ticket subjects or technical details. The goal is to make the member feel heard, not to troubleshoot. #### Voice Message Content When using the Voice channel, the system: * Asks if now is a good time to speak * Confirms speaking with the right person before discussing details * Expresses genuine concern about recent experience * Asks open-ended questions about how things are going * Does NOT reference specific ticket numbers, technical details, or individual issues * Offers to arrange a follow-up with the team if concerns remain unresolved * Detects voicemails and hangs up immediately without leaving a message #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | ---------------------------------------------------------------- | | **Cooldown** | 168 hours (1 week) | Prevents duplicate outreach to same member | | **Max Actions/Day** | 5–10 | Requires thoughtful review and personalization | | **Auto-Execute** | ❌ Disabled | Always review context before sending | | **Confidence Threshold** | 70–95% | Auto-calculated based on ticket count, priority, and open status | #### Confidence Scoring Confidence is automatically calculated based on: * **Base confidence:** 70% (threshold met) * **+5%** for each ticket above the minimum (up to +15%) * **+5%** if any tickets are High or Critical priority * **+3%** if member has open (unresolved) tickets * **Maximum:** 95% #### Priority Determination Priority is automatically determined based on: * **High:** Ticket count ≥ 2x threshold OR any Critical-priority tickets * **Medium:** Default for standard patterns #### Available Variables Use these in custom prompts or voice messages: * `{{customer_name}}` — Full name of the customer * `{{customer_email}}` — Customer email address * `{{ticket_count}}` — Number of recent support tickets * `{{lookback_days}}` — Configured lookback window in days * `{{company_name}}` — Name of your space/location (voice messages) * `{{session_id}}` — Unique session identifier (voice messages) #### Best Practices * Always review before sending — understand the context and ticket severity * Escalate to your most senior support person or manager for high-priority cases * Consider a phone call instead of email for critical patterns or repeated issues * Track whether your outreach resolves the underlying issue and reduces ticket volume * Adjust the Minimum Ticket Count and Lookback Window based on your space's support volume * Use voice outreach for high-value members or critical situations *** ### 🟢 Support Category Spike **Status:** ✅ Live\ **Category:** Support\ **Channel:** Internal note only\ **Auto-Execute Safe?** ✅ Yes (Internal only) #### What It Does Detects statistically significant spikes in support issues within specific categories compared to a rolling average baseline. When a spike is detected, it creates an internal alert for operators—not an outbound message to customers—to notify them of potential systemic issues requiring attention. Support tickets are automatically classified into categories using AI analysis (Printing, WiFi, Access, Billing, Noise, HVAC, Cleaning, Booking, or Other). The agent then compares current weekly ticket volume against the historical average to identify anomalies. **Example alert:** "Printing issues increased 300% this week (12 tickets vs. average 3). Possible printer problem." #### When It Triggers * A support category's weekly ticket count exceeds the rolling average by the configured spike multiplier (default: 2x increase) * The historical baseline has at least the minimum number of tickets (default: 3) to ensure statistical significance * Only formal support tickets are analyzed (tickets handled entirely by AI chat are excluded) * One alert per category per week (deduplication prevents duplicate alerts) #### What It Monitors * Support ticket categories (auto-classified by AI on first evaluation) * Current week's ticket count per category * Rolling average baseline (configurable: 2–12 weeks, default 4 weeks) * Spike magnitude (ratio of current week vs. baseline average) * Ticket classification accuracy and confidence scores #### Configuration Parameters | Parameter | Default | Description | Range | | ---------------------------- | ------- | ---------------------------------------------------------------------------------------------------------- | ------------ | | **Baseline Weeks** | `4` | Number of historical weeks to calculate the rolling average baseline | 2–12 weeks | | **Spike Multiplier** | `2` | How many times above the baseline average to trigger an alert (e.g., 2x means 200% increase) | 1.5–5x | | **Minimum Baseline Tickets** | `3` | Minimum average tickets per week in baseline to trigger alerts (prevents noise from low-volume categories) | 1–10 tickets | #### Message Content Creates an internal note for operators with: * Category name (e.g., "WiFi Access Issues") * Current week's ticket count * Historical baseline average for comparison * Percentage increase (e.g., "300% increase") * Sample ticket subjects from the spike period (top 3 most recent) * Confidence score based on statistical significance * Priority level (High for large spikes, Medium for moderate increases) #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | -------------------------------------------- | | **Cooldown** | 168 hours (1 week) | One alert per category per week | | **Max Actions/Day** | 3–5 | Prevents alert fatigue across all categories | | **Auto-Execute** | ✅ Enabled | Internal notes are safe to auto-create | | **Confidence Threshold** | N/A | Not applicable for internal notes | #### Available Variables Use these in internal note templates: * `{{category}}` — Support issue category name (Printing, WiFi, Access, Billing, Noise, HVAC, Cleaning, Booking, or Other) * `{{ticket_count}}` — Number of tickets in current week for this category * `{{baseline_average}}` — Historical average tickets per week (rounded to nearest whole number) * `{{percentage_increase}}` — Spike magnitude as percentage (e.g., "300%") * `{{spike_multiplier}}` — Current week as multiple of baseline (e.g., "4x") * `{{baseline_weeks}}` — Number of weeks used for baseline calculation * `{{sample_subjects}}` — Top 3 most recent ticket subjects in this category * `{{confidence_score}}` — Statistical confidence score (0–100%) * `{{priority_level}}` — Alert priority (High or Medium) #### Support Issue Categories The AI classifier automatically categorizes support tickets into these 9 categories: | Category | Examples | | ------------ | --------------------------------------------------------------------------------------------- | | **Printing** | Printers, print jobs, print credit, scanners, copiers, paper jams, toner, print quality | | **WiFi** | Internet connectivity, WiFi password, slow internet, network access, VPN, ethernet | | **Access** | Building access, key cards, door codes, access control, entry/exit issues, badge problems | | **Billing** | Invoice questions, payment issues, subscription problems, refunds, pricing, charges | | **Noise** | Noise complaints, loud members, phone booth availability, quiet area violations, music volume | | **HVAC** | Temperature issues, AC problems, heating, ventilation, air quality | | **Cleaning** | Dirty spaces, trash collection, spills, restroom cleanliness, desk maintenance, kitchen mess | | **Booking** | Meeting room bookings, calendar conflicts, cancellations, availability, desk reservations | | **Other** | General inquiries, suggestions, events, amenities, parking, mail/delivery, coffee/supplies | #### Best Practices * Enable immediately for all locations with support ticket volume * Review alerts daily and assign to appropriate team members (e.g., IT for WiFi or Printing spikes) * Track whether alerts correlate with infrastructure issues or maintenance needs * Use spike patterns to schedule proactive maintenance before customer complaints increase * Adjust **Baseline Weeks** based on your space's seasonality (longer baselines for more stable averages) * Lower **Minimum Baseline Tickets** if you want alerts for lower-volume categories * Consider increasing **Spike Multiplier** if you receive too many alerts (e.g., from 2x to 3x) * Review classification accuracy periodically — AI can be retrained if categories need refinement * For multi-location businesses, alerts are business-wide; monitor per-location trends in your analytics dashboard *** ## Knowledge Agents ### 🟢 FAQ Gap Detection **Status:** ✅ Live\ **Category:** Knowledge\ **Channel:** Internal note only\ **Auto-Execute Safe?** ✅ Yes (Internal only) #### What It Does Detects recurring questions that members ask but have no FAQ article to answer. When the same question theme is asked multiple times across different conversations within a configurable window, it creates an internal note for operators proposing a new FAQ article with an AI-generated draft based on available business information. Rather than simply alerting you to gaps, this agent composes a ready-to-review FAQ draft using your space's resources, plans, opening hours, and previously answered support tickets as context. **Example output:** "Members are asking about guest day-pass pricing (asked 8 times this week). Suggested FAQ draft attached." #### When It Triggers * The same question theme is asked N+ times (default: 3 times) within a configurable lookback window (default: 7 days) * Questions are collected from two sources: * **HelpDesk auto-reply failures** — Where the AI could not find a matching FAQ to auto-reply * **AI Chat FAQ tool failures** — Where the interactive chat's FAQ lookup returned no relevant article * Questions are clustered by semantic similarity (e.g., "What time do you open?", "Opening hours?", "When does the space close?" are grouped together) * One internal note per question theme per evaluation cycle (deduplication prevents duplicate suggestions) #### What It Monitors * HelpDesk messages where AI auto-reply failed (FAQ embedding match below threshold) * AI Chat sessions where the FAQ tool returned no relevant article * Question semantic similarity across all unanswered questions * Frequency of each question theme within the lookback window #### Configuration Parameters | Parameter | Default | Description | Range | | ---------------------------- | ------- | ------------------------------------------------------------------------------------ | ---------------- | | **Minimum Occurrences** | `2` | Number of times a question theme must be asked to trigger a suggestion | 1–50 occurrences | | **Lookback Window** | `21` | Number of days to look back for unanswered questions | 1–90 days | | **Similarity Threshold** | `0.70` | How similar questions must be to be grouped together (higher = stricter matching) | 0.50–0.95 | | **Max Questions to Process** | `100` | Maximum number of unanswered questions to analyze per evaluation (most recent first) | 10–500 | #### Message Content Creates an internal note for operators with: * **Question theme** — Normalized label for the recurring question (e.g., "Guest day-pass pricing") * **Occurrence count** — How many times this question was asked in the lookback window * **Sample questions** — Up to 5 actual questions from real conversations showing how members phrased the question * **AI-generated FAQ draft** including: * **Title** — Clear, searchable question title (e.g., "What are the opening hours?") * **Summary** — 1–2 sentence answer * **Full Text** — Detailed answer using business context (resources, plans, opening hours, etc.) * **Context sources used** — Which business data was used to compose the draft (e.g., "Used opening hours, resource list, and 2 previously answered support tickets") * **"Needs Operator Input" markers** — Sections marked `[NEEDS OPERATOR INPUT]` where the AI didn't have enough context to answer accurately * **"Create FAQ" action button** — Routes to the FAQ editor with pre-filled title and body for quick review and publishing #### Draft Generation Context The AI draft is composed using the following business information: | Context Source | Data Used | Why | | --------------------------------------- | -------------------------------------- | ---------------------------------------------------- | | **Business details** | Name, opening hours, address, timezone | Many questions relate to basic location info | | **Resources** | Resource names, descriptions, types | "How do I book the meeting room?" needs room names | | **Plans** | Plan names, descriptions, prices | Pricing and plan questions are common | | **Previously answered support tickets** | Human answers to similar questions | Excellent source material for FAQ content | | **Sample unanswered questions** | Actual member phrasing (top 5) | Ensures the draft addresses how members actually ask | #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | ------------------------------------------------------------ | | **Cooldown** | 168 hours (1 week) | Batch review of content gaps; prevents duplicate suggestions | | **Max Actions/Day** | 3–5 | Manageable for content team review | | **Auto-Execute** | ✅ Enabled | Internal notes create no customer impact | | **Confidence Threshold** | N/A | Not applicable for internal notes | #### Available Variables Use these in internal note templates: * `{{question_theme}}` — Normalized topic label (e.g., "Guest day-pass pricing") * `{{occurrence_count}}` — Number of times this question was asked * `{{lookback_days}}` — Configured lookback window in days * `{{sample_questions}}` — Up to 5 actual questions from real conversations * `{{draft_title}}` — AI-generated FAQ title * `{{draft_summary}}` — AI-generated 1–2 sentence summary * `{{draft_full_text}}` — AI-generated full FAQ article text * `{{needs_operator_input}}` — Boolean indicating if the draft has sections needing human input * `{{context_sources_used}}` — List of business data sources used for the draft * `{{similarity_threshold}}` — Configured similarity threshold for clustering #### Best Practices * Enable immediately for all locations with AI chat or HelpDesk auto-reply enabled * Review detected gaps weekly as a batch rather than individually * Prioritize by frequency — most-asked questions should become FAQs first * Assign FAQ creation to your content manager or knowledge owner * Use the "Create FAQ" action button to publish drafts quickly after review * Track FAQ implementation and remeasure after 2 weeks to verify the gap is closed * Measure reduction in unanswered questions as your success metric * Adjust **Minimum Occurrences** based on your space's question volume (lower for busier spaces) * Lower **Similarity Threshold** if you want broader question grouping (e.g., 0.75 instead of 0.80) * Increase **Lookback Window** if you want to detect slower-building patterns (e.g., 14 days instead of 7) * Review `[NEEDS OPERATOR INPUT]` sections and fill in missing information before publishing *** ## Sales Agents ### 🔜 Lead Follow-up **Status:** 🔜 Coming Soon\ **Category:** Sales\ **Channel:** Email\ **Auto-Execute Safe?** ⚠️ With caution #### What It Does Proposes follow-up messages for potential leads who inquired but haven't converted. Monitors AI sessions from non-members who asked about tours, pricing, or availability. #### When It Triggers * Lead completed AI session asking about tours, pricing, or plans * No conversion occurred (no booking or membership signup) * Configurable dormancy period has passed (e.g., 7 days) * Lead hasn't already received recent outreach #### What It Monitors * AI conversation intent from non-members * Engagement level in AI chat * Topics discussed (pricing, availability, tour interest) * Time since last interaction * Conversion status #### Message Content Follow-up includes: * Reference to their earlier inquiry * Personalized response to their specific questions * Clear call-to-action (schedule tour, get pricing, start trial) * Offer to answer remaining questions * Time-limited incentive (if configured) #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | ------------------------------------ | | **Cooldown** | 168 hours (1 week) | Respects lead fatigue | | **Max Actions/Day** | 10–20 | Scales with lead flow | | **Auto-Execute** | ⚠️ Use caution | Safe once you've validated messaging | | **Confidence Threshold** | 80%+ | Only for high-intent leads | #### Available Variables * `{{lead_name}}` — Lead's full name * `{{inquiry_topic}}` — What they asked about (tour, pricing, etc.) * `{{inquiry_date}}` — When they inquired * `{{inquiry_details}}` — Specific questions they asked * `{{available_tours}}` — Upcoming tour dates * `{{pricing_link}}` — Link to pricing page #### Best Practices * Start with manual review — validate that message quality is good * Personalize responses based on what they actually asked * Don't enable auto-execute until you've reviewed 20+ proposed messages * Monitor conversion rates to leads you follow up with * Segment leads by intent level (tour interest vs. general inquiry) * Adjust timing based on sales cycle length *** ### 🔜 Tour Reminders **Status:** 🔜 Coming Soon\ **Category:** Sales\ **Channel:** Email\ **Auto-Execute Safe?** ✅ Yes #### What It Does Sends friendly reminders before scheduled tours with helpful preparation information. Monitors upcoming tour bookings and proposes reminder messages at configurable intervals. #### When It Triggers * Tour is scheduled for future date * Reminder timing configured (e.g., 24 hours before, 1 hour before) * Tour hasn't been cancelled * Previous reminder hasn't already been sent #### What It Monitors * Tour booking schedules * Prospect contact information * Tour timing and location details * Tour cancellation status #### Message Content Reminders include: * Personalized greeting to the prospect * Tour date, time, and location * Parking information and directions * What to bring (ID, questions to prepare) * Building access instructions * Contact number if they need to reschedule * Optional: virtual tour link if they can't attend in person #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------- | ------------------------------- | | **Cooldown** | 24 hours | One reminder per day | | **Max Actions/Day** | 20–50 | Scales with tour volume | | **Auto-Execute** | ✅ Enabled | Helpful, low-risk communication | | **Confidence Threshold** | 85%+ | High confidence in tour details | #### Available Variables * `{{prospect_name}}` — Tour attendee's name * `{{tour_time}}` — Scheduled date and time * `{{tour_location}}` — Address and directions * `{{parking_info}}` — Parking instructions * `{{building_access}}` — How to enter the building * `{{reschedule_link}}` — Link to reschedule if needed * `{{contact_number}}` — Your phone number * `{{location_name}}` — Name of your space #### Best Practices * Enable immediately — tour reminders are universally appreciated * Send first reminder 24 hours before tour * Consider second reminder 1 hour before for same-day confirmation * Include parking and access details to reduce no-shows * Monitor tour attendance and no-show rates * Adjust reminder timing based on your show-up rate *** ### 🔜 Tour No-Show Follow-up **Status:** 🔜 Coming Soon\ **Category:** Sales\ **Channel:** Email\ **Auto-Execute Safe?** ✅ Yes #### What It Does Reaches out to prospects who scheduled a tour but did not show up. Detects when a tour booking occurred but the prospect didn't check in or arrive, and proposes a friendly follow-up. #### When It Triggers * Tour booking was scheduled * Tour time has passed * No check-in or arrival was recorded * Booking is marked as "no-show" #### What It Monitors * Tour booking schedules vs. actual attendance * Check-in records * Prospect contact information * No-show status #### Message Content Follow-up includes: * Friendly, non-judgmental tone * Acknowledgment of missed tour * Offer to reschedule with specific available times * Question about what might have prevented attendance * Assurance of willingness to accommodate their schedule * Alternative: virtual tour option #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------- | --------------------------------- | | **Cooldown** | 24 hours | Send day-of or day-after | | **Max Actions/Day** | 10–20 | Depends on no-show rate | | **Auto-Execute** | ✅ Enabled | Helpful follow-up, safe messaging | | **Confidence Threshold** | 90%+ | Only for confirmed no-shows | #### Available Variables * `{{prospect_name}}` — Name of prospect who missed tour * `{{tour_date}}` — Date they were scheduled for * `{{tour_time}}` — Time of missed tour * `{{available_times}}` — Next available tour times * `{{reschedule_link}}` — Link to reschedule * `{{contact_number}}` — Your phone number * `{{virtual_tour_link}}` — Link to virtual tour if available * `{{location_name}}` — Name of your space #### Best Practices * Enable immediately — follow-up quickly after no-show * Keep tone friendly and understanding * Offer multiple rescheduling options (dates and times) * Include virtual tour link as alternative * Track reschedule conversion rate * Consider reaching out via phone call for high-value leads *** ### 🔜 Lead Qualification Scoring **Status:** 🔜 Coming Soon\ **Category:** Sales\ **Channel:** Internal note only\ **Auto-Execute Safe?** ✅ N/A (Internal only) #### What It Does Scores and qualifies leads based on conversation interactions and enrichment data. Analyzes AI chat sessions with prospects to extract qualification signals and determine sales priority. #### When It Triggers * Lead completes an AI conversation (or reaches certain length/engagement) * Sufficient signals extracted to score lead quality * Score changes lead's tier (hot → warm, warm → cold, etc.) * Integration with CRM systems for pipeline sync #### What It Monitors * AI chat session length and engagement * Budget mentions and price sensitivity * Team size and space requirements * Move-in timeline and urgency * Must-have features and nice-to-haves * Decision-maker presence * Objection handling #### Message Content Creates internal note with: * Lead qualification score (Hot/Warm/Cold) * Key signals and reasoning * Recommended next action * Suggested follow-up timing * CRM sync status * Priority assignment for sales team #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------- | ----------------------------------- | | **Cooldown** | 24 hours | Re-score only once per day | | **Max Actions/Day** | 50+ | Can handle volume of prospects | | **Auto-Execute** | ✅ Enabled | Internal notes, auto-create is safe | | **Confidence Threshold** | N/A | Not applicable | #### Available Variables * `{{prospect_name}}` — Prospect's name * `{{score_tier}}` — Hot/Warm/Cold * `{{confidence_score}}` — Numeric confidence (0-100%) * `{{key_signals}}` — Main qualifying factors * `{{recommended_action}}` — Next step (call, follow-up, etc.) * `{{suggested_timeline}}` — When to follow up #### Best Practices * Enable to segment leads by priority * Set up CRM integration for automatic pipeline sync * Use Hot tier for immediate sales outreach * Monitor scoring accuracy — adjust signals quarterly * Train sales team on how to interpret scoring signals * Track conversion by lead tier to validate scoring model *** ## Engagement Agents ### 🔜 Onboarding Nudges **Status:** 🔜 Coming Soon\ **Category:** Engagement\ **Channel:** Email\ **Auto-Execute Safe?** ✅ Yes #### What It Does Sends helpful reminders when new customers haven't completed expected first-week actions. Tracks onboarding milestones for new members and proposes nudge messages when expected actions haven't been completed. #### When It Triggers * New customer joins your space * First-week milestone actions are tracked: * First booking not made within 3 days * WiFi not connected within 2 days * App not downloaded within 3 days * Welcome email not opened within 1 day * Customizable thresholds for each milestone #### What It Monitors * Customer account creation date * First booking date * App downloads and logins * WiFi network connections * Email opens and clicks * Member portal access #### Message Content Nudges include: * Friendly, encouraging tone * What milestone is missing * Why it matters (benefits of completing) * Step-by-step instructions to complete * Link or direct path to take action * Offer to help if stuck #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------- | ----------------------------------- | | **Cooldown** | 24 hours | One nudge per day per milestone | | **Max Actions/Day** | 10–20 | Scales with new member rate | | **Auto-Execute** | ✅ Enabled | Helpful, low-risk engagement | | **Confidence Threshold** | 85%+ | High confidence in onboarding needs | #### Available Variables * `{{customer_name}}` — New member's first name * `{{signup_date}}` — When they joined * `{{missing_milestone}}` — What they haven't done yet * `{{milestone_benefit}}` — Why they should do it * `{{action_link}}` — Direct link to complete action * `{{phone_number}}` — Your phone number * `{{location_name}}` — Name of your space #### Best Practices * Enable immediately for all new customers * Customize milestones based on your onboarding flow * Keep instructions short and clear * Link directly to actions (WiFi setup, app store, etc.) * Track completion rates by milestone * Consider phone check-in for VIP or corporate memberships * Adjust thresholds based on your space's typical onboarding timeline *** ### 🔜 Booking No-Show Follow-up **Status:** 🔜 Coming Soon\ **Category:** Engagement\ **Channel:** Email\ **Auto-Execute Safe?** ✅ Yes #### What It Does Follows up with customers who booked a resource (meeting room, desk, etc.) but did not check in or show up. Monitors bookings and detects no-shows, then proposes a check-in message. #### When It Triggers * Resource booking is scheduled (past date/time) * No check-in or arrival recorded * Booking is marked as "no-show" * Previous follow-up hasn't already been sent #### What It Monitors * Resource booking schedules * Check-in records * Customer history of no-shows * Booking cancellation vs. no-show pattern #### Message Content Follow-up includes: * Friendly check-in tone * Acknowledgment of missed booking * Confirmation that space is still available if rescheduled * Reminder about cancellation policies * Option to reschedule for future date * Offer of assistance with future bookings #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ----------------- | --------------------------------- | | **Cooldown** | 48 hours | Not too aggressive | | **Max Actions/Day** | 10–15 | Depends on no-show rate | | **Auto-Execute** | ✅ Enabled | Helpful, low-risk follow-up | | **Confidence Threshold** | 85%+ | High confidence in no-show status | #### Available Variables * `{{customer_name}}` — Member's name * `{{resource_name}}` — Room or desk booked * `{{booking_date}}` — Date of booking * `{{booking_time}}` — Time of booking * `{{cancellation_policy}}` — Your cancellation terms * `{{reschedule_link}}` — Link to reschedule * `{{contact_number}}` — Your phone number * `{{location_name}}` — Name of your space #### Best Practices * Enable immediately — follow-up quickly after no-show * Use friendly, understanding tone * Include clear cancellation policy reminders * Make rescheduling easy with direct link * Track no-show patterns by customer * Consider charging cancellation fee only after pattern of no-shows * Proactively reach out to repeat offenders via phone *** ## Retention Agents ### 🔜 At-Risk Member Detection **Status:** 🔜 Coming Soon\ **Category:** Retention\ **Channel:** Email\ **Auto-Execute Safe?** ❌ No #### What It Does Identifies customers showing declining engagement and proposes proactive retention outreach. Analyzes usage patterns and detects statistically significant drops compared to historical baseline. #### When It Triggers * Usage metrics show significant decline: * Check-in frequency down 50%+ from previous period * Booking frequency down significantly * Payment patterns change (late payments, disputes) * Change is sustained over configurable period (e.g., 2 weeks) * Customer is not already flagged for other retention efforts * No recent proactive outreach attempted #### What It Monitors * Historical check-in patterns * Booking frequency trends * Invoice payment history * Resource usage intensity * Time since last check-in * Contract approaching expiry #### Message Content Retention outreach includes: * Personalized acknowledgment of customer * Specific observation about their changing usage * Offer of help to understand barriers * Highlighting unused benefits or features * Suggestion to adjust plan if needs have changed * Value reminder and company culture fit * Direct access to manager for one-on-one discussion #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | -------------------------------- | | **Cooldown** | 168 hours (1 week) | Space out retention efforts | | **Max Actions/Day** | 5–10 | Requires thoughtful review | | **Auto-Execute** | ❌ Disabled | Always requires personal context | | **Confidence Threshold** | N/A | Manual review only | #### Available Variables * `{{customer_name}}` — Member's name * `{{previous_avg_checkins}}` — Historical check-in rate * `{{current_checkins}}` — Recent check-in rate * `{{usage_change}}` — Percentage decline * `{{days_since_checkin}}` — Days inactive * `{{unused_features}}` — Benefits they don't use * `{{plan_name}}` — Current plan * `{{manager_name}}` — Manager's name for personal outreach * `{{location_name}}` — Name of your space #### Best Practices * Never auto-execute — always review for context * Personalize based on their specific usage pattern * Highlight specific benefits they're not using * Offer plan adjustment or flexibility * Consider offering temporary rate discount or waived fees * Escalate to manager for high-value customers * Track whether retention efforts successfully re-engage customer *** ### 🔜 Cancellation Risk Detection **Status:** 🔜 Coming Soon\ **Category:** Retention\ **Channel:** Internal note only\ **Auto-Execute Safe?** ✅ N/A (Internal only) #### What It Does Identifies early warning signals that a customer might cancel their contract. Uses multiple signals to predict cancellation risk and creates internal alerts for high-risk members. #### When It Triggers * Multiple risk signals detected (weighted scoring): * Declining usage (50 points) * Overdue payment (40 points) * Support ticket sentiment negative (30 points) * Contract approaching end date (20 points) * Failed payment methods (30 points) * Total risk score exceeds threshold (e.g., 75 points = high risk) * Risk is flagged for internal action #### What It Monitors * Check-in trends and patterns * Invoice payment status * Support ticket sentiment and frequency * Contract end date proximity * Payment method status * Engagement in location events * NPS/survey scores if available #### Message Content Creates internal alert with: * Customer name and contract details * Risk score and severity (Low/Medium/High) * Specific risk signals triggered * Days until contract renewal opportunity * Suggested retention actions * Recommended next step (call, offer, etc.) #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | ----------------------------------- | | **Cooldown** | 168 hours (1 week) | Weekly risk assessment | | **Max Actions/Day** | 5–10 | Depends on customer base size | | **Auto-Execute** | ✅ Enabled | Internal notes, auto-create is safe | | **Confidence Threshold** | N/A | Not applicable | #### Available Variables * `{{customer_name}}` — Member's name * `{{risk_score}}` — Numeric risk (0-100) | * `{{risk_tier}}` — Low/Medium/High * `{{risk_factors}}` — List of triggers * `{{days_to_expiry}}` — Days until renewal * `{{recommended_action}}` — Suggested next step * `{{manager_assignment}}` — Recommended manager to handle #### Best Practices * Review high-risk members immediately * Assign to manager for proactive outreach * Don't wait until contract renewal deadline * Offer preventative incentives before ask for retention * Track which interventions are most effective * Monitor risk score trends over time * Celebrate successes when at-risk customer re-engages *** ## Revenue Agents ### 🔜 Upsell Opportunities **Status:** 🔜 Coming Soon\ **Category:** Revenue\ **Channel:** Email\ **Auto-Execute Safe?** ❌ No #### What It Does Detects when customer usage patterns suggest they would benefit from an upgrade. Monitors usage intensity and frequency to identify customers who consistently need more than their current plan offers. #### When It Triggers * Customer usage exceeds plan limits: * Frequently books last-minute resources (capacity constraint) * Day passes purchased alongside main plan (coverage gap) * Extra hour credits purchased frequently (time gap) * Meeting room requests denied due to availability * Pattern sustained over configurable period (e.g., 30 days) * Upgrade plan is relevant to their usage pattern * Customer not recently offered upsell #### What It Monitors * Booking frequency relative to plan allowance * Day pass and credit purchases * Resource availability and booking denials * Usage growth trends * Customer segment and plan tier * Contract tenure and growth trajectory #### Message Content Upsell opportunity includes: * Data-driven observation: "You've booked a meeting room 15 times in the past month, but your plan includes 5. We noticed you're consistently running out..." * Specific upgrade recommendation with ROI * Comparison of current plan vs. recommended plan * Cost savings from consolidating separate purchases * Trial period offer if new plan type * Success story from similar customer * Limited-time upgrade incentive #### Recommended Settings | Setting | Recommended Value | Why | | ------------------------ | ------------------ | ----------------------------- | | **Cooldown** | 168 hours (1 week) | Don't be aggressive | | **Max Actions/Day** | 5–10 | Requires personalization | | **Auto-Execute** | ❌ Disabled | Sensitive to customer context | | **Confidence Threshold** | N/A | Manual review recommended | #### Available Variables * `{{customer_name}}` — Member's name * `{{current_plan}}` — Their current plan * `{{recommended_plan}}` — Suggested upgrade * `{{usage_metric}}` — What exceeded (bookings, credits, etc.) * `{{usage_amount}}` — How much they used * `{{plan_limit}}` — Plan limit exceeded * `{{roi_savings}}` — Monthly savings if upgraded * `{{upgrade_link}}` — Link to upgrade * `{{trial_period}}` — Any trial offer #### Best Practices * Never auto-execute — upsells need personalization * Start with data (show their actual usage) * Lead with value, not price * Offer trial period for new plan types * Consider proactive discount for price-sensitive customers * Group related upsells together * Track conversion rate by segment * Monitor churn after rejected upsells (might indicate pricing issue) *** ## Configuration Summary by Category ### Safe for Auto-Execute These agents are good candidates for auto-execute after initial testing: * ✅ **Due Invoice Reminders** — Factual, expected * ✅ **Contract Renewal** — Relationship-building, timely * ✅ **Onboarding Nudges** — Helpful, low-risk * ✅ **Tour Reminders** — Expected, appreciated * ✅ **Tour No-Show Follow-up** — Friendly, helpful * ✅ **Booking No-Show Follow-up** — Gentle check-in ### Requires Manual Review These agents benefit from human judgment before sending: * ❌ **Support Pattern Detection** — Needs context understanding * ❌ **At-Risk Member Detection** — Personal touch needed * ❌ **Upsell Opportunities** — Sensitive to customer relationship * ⚠️ **Lead Follow-up** — Quality control important * 📝 **Internal agents** — Review before action taken ### Internal Notes Only These create notes for your team, not customer messages: * 📝 **Support Category Spike** * 📝 **FAQ Gap Detection** * 📝 **Cancellation Risk Detection** * 📝 **Lead Qualification Scoring** *** ## Getting Started with Agents ### Week 1: Foundation 1. Enable **Due Invoice Reminders** and **Contract Renewal** 2. Keep both on manual review for first 10 actions each 3. Review AI Inbox daily to understand agent quality 4. Monitor customer responses and feedback ### Week 2–3: Expansion 1. If satisfied with quality, enable auto-execute at 85% threshold 2. Monitor approval rates and adjust confidence threshold as needed 3. Consider enabling **Tour Reminders** if you conduct tours 4. Keep other agents on manual review for learning ### Week 4+: Optimization 1. Analyze approval rates and rejection patterns 2. Adjust cooldowns and rate limits based on your workflow 3. Consider expanding to engagement agents (Onboarding Nudges, etc.) 4. Build custom prompts to match your location's voice *** ## Related Documentation * [Configuring Agents](/platform/ai/configuring-agents) — Step-by-step configuration guide * [Proactive Agents Overview](/platform/ai/proactive-agents) — How the system works * [AI Inbox](/platform/ai/inbox) — Review and approve proposed actions * [AI Conversations](/platform/ai/ai-conversations) — View two-way conversations