Platform/Pulse AI
Featured
Pulse AI

Speak hotel.
Get answers.

Pulse — your AI Revenue Manager. Built for hotel data, with private memory for your hotel, measurable accuracy, and guardrails you can audit.

No prompt engineering. No black box. Ask in plain language; Pulse runs the report, draws the chart, and explains the answer.

app.peaqplus.com/pulse
HISTORY
Pickup last 30d
Weekend gap
Compset Aug
Daily Briefing
How has pickup changed in the last 30 days?
P
Pickup is up +9.4% versus the same point last year. The acceleration started 14 days ago, driven mostly by direct bookings (+18%) and short-lead corporate (+11%).
Compare to budgetDrill into corporateShow channel mix
Pulse is ready · ⌘K to query
01
Pulse Chat

Natural-language revenue analysis. With charts.

Ask Pulse anything about your hotel — "How did our weekend pickup compare to last year?", "Which segment is driving the booking pace shift?", "Show me the booking curve for next month."

Pulse picks the right tool from its kit, runs the query against your live PMS data, and answers in your language with a table, a chart, three suggested next-step actions, and a deep-link to the source report.

What makes it different
  • Runs against your snapshot data — same source as your reports
  • Multi-turn conversation; the AI remembers what you asked two messages ago
  • Charts rendered inline, with dual-axis where the comparison needs it
  • Suggested actions — three one-click follow-up questions on every answer
  • "Go to the report" deep-link — every tool answer carries a direct link
  • Voice in, voice out — speech recognition in your language
  • Real-time delivery — answers arrive instantly, with a reliable fallback if the connection drops
  • Conversation auto-deactivates when fresh PMS data lands — no stale-data answers
app.peaqplus.com/pulse
HISTORY
Pickup last 30d
Weekend gap
Compset Aug
Daily Briefing
How has pickup changed in the last 30 days?
P
Pickup is up +9.4% versus the same point last year. The acceleration started 14 days ago, driven mostly by direct bookings (+18%) and short-lead corporate (+11%).
Compare to budgetDrill into corporateShow channel mix
Pulse is ready · ⌘K to query
02
What you can ask

Plain-language questions. Real-time answers.

Pulse never just talks — every numerical answer comes from a query against your live data. The user never picks a tool; Pulse picks. The result has the same shape every time: a chart, a table, and three suggested next-step questions you can click.

The kit grows as we add new questions Pulse can handle — but the user-facing surface stays the same. You ask. Pulse answers.

What you can ask about
  • Performance & pace — pickup, segments, budget tracking, YoY same-point
  • Forward look — booking curves, weak days, monthly trajectory
  • Anomalies — ADR outliers, unusual ratios, channel surprises
  • Operational decisions — where to push, what to investigate, what changed
EXAMPLE QUESTIONS
QHow did last weekend's pickup compare to last year same point?
ALast weekend: 18 room nights picked up. Same point last year: 24. Gap −6 (−25%). Corporate segment is the underperformer — −8 vs LY on the Saturday alone.
Δ vs LY−25%
QHow are we tracking vs Q2 plan?
AQ2 pace +3.2% vs LY. Room-night gap to budget: −340. Required daily pickup to hit budget: 18/day. Current run-rate: 14/day. Three weak weekdays in the next 21 days are dragging the gap.
Required / actual18 / 14 per day
QAny unusual ADR moves in the last 7 days?
ATwo days flagged: May 14 (€142 vs €178 7-day average, −20%) and May 16 (€201 vs €178, +13%). The May 14 dip aligns with a large OTA discount push.
Outliers2 days
QWhich days next month need a marketing push?
AThree weak days in the next 21d: May 12 (38% projected occ), May 14 (42%), May 18 (44%). Suggested next step: "Draft ads for these dates" — one click.
Days flagged3 of 21
Every answer comes with a chart, a table, and three one-click follow-up questions.
03
The agents

Three personalities. One platform.

Every new conversation starts by picking an agent — a specialised AI personality with its own focus and system instruction. The Pickup Analyst dives into booking dynamics. The Forecast Advisor reads the next 60 days with event context. The Pricing Expert thinks in ADR / RevPAR / rate-volume tradeoffs. The agents share the same toolkit and the same data; what differs is the lens and the prompt.

Language follows the user — chosen per-account. The agent's persona is consistent across conversations.

Agents available
  • Pickup Analyst — booking dynamics, daily and weekly variation (live)
  • Forecast Advisor — 60-day projection with event context (live)
  • Pricing Expert — ADR, RevPAR, rate-volume strategy (coming soon)
NEW CONVERSATION · CHOOSE AGENT
Pick the lens that fits the question.
📈
Pickup AnalystLive
Booking dynamics. Pickup analysis. Daily and weekly variation.
Start →
🔮
Forecast AdvisorLive
60-day projection with event context. Accuracy-trend reads.
💎
Pricing ExpertComing soon
ADR / RevPAR / rate-volume strategy. Pricing tradeoffs.
All agents share the toolkit · language follows your account
04
Tell Pulse about your hotel

Strategy-aware AI. With the hotel name stripped.

Pulse learns from observation — but you can also tell it about your hotel directly. Three free-text fields on the AI Settings page describe your pricing strategy, your market context, and any special circumstances right now. Pulse uses these in every prompt to read the same numbers correctly: a 72% occupancy isn't an alarm if your strategy is yield-focused.

The hotel name itself is stripped from the context before any prompt is built — and an AI validator checks every save to confirm you haven't included it accidentally. The model never identifies the specific property; it only sees the strategic shape of the business.

The three fields
  • Sales strategy — yield · balanced · volume (changes what counts as a problem)
  • Market context — free text · "Pulse Suggest" button auto-drafts based on hotel type + location
  • Special circumstances — free text · "Renovation on floor 3 until May", "Group block Mar 18–22"
  • Hotel-name privacy — stripped before prompt, AI-validated on save · bias and privacy guard
Security & isolation
app.peaqplus.com/pulse-chat/settings
AI SETTINGS · HOTEL CONTEXT
Tell Pulse about your hotel.
Sales strategy
Yield
higher ADR, lower occ ok
Balanced
both matter
Volume
fill the rooms
Market context✨ Pulse Suggest
4-star boutique hotel in the city center. Source markets: Czech and Polish families (long weekends), business travellers (Tue–Thu). Competitive set: 5 design-led 4-star properties within 1 km.
Special circumstances
Renovation on the 3rd floor until May 15. 12 rooms out of inventory through Apr.
Privacy enforced: hotel name is stripped before any prompt · AI validator checks every save.
05
Memory

Pulse learns your hotel. Yours alone.

Most AI tools start every conversation from zero. Pulse builds memory — the patterns, the context, the things your team has confirmed are true — kept private to your hotel.

Every third message, Pulse extracts candidate memories from the conversation and surfaces them for review. Three confidence levels: Pending (AI-proposed, awaiting confirmation), Active (multi-conversation-confirmed), Manual (admin-added). The admin can edit or delete any of them; the AI Settings page is the single edit surface.

What gets remembered
  • Hotel-level — sales strategy, market context, source markets, segment patterns
  • Per-user — query patterns, language preference, dashboard usage
  • Auto-learning — every 3rd message proposes new memories; admin confirms
  • Three confidence levels — Pending (0.4) · Active (0.7) · Manual (1.0)
  • Other hotels' data cannot reach yours — the boundary lives in the data layer, not in a policy you trust
YOUR HOTEL · ISOLATED BOUNDARY
1.0
Manual
Explicit team-added context
0.7
Active
Confirmed across conversations
0.4
Pending
Newly observed, under observation
Every memory stays with your hotel.
Other hotels' data cannot reach yours — by design.
06
Forecast you can verify

Most AI forecasts promise. Pulse measures.

Smart Forecast Enhanced corrects your daily 60-day pickup forecast using a frontier reasoning model, calibrated on a per-hotel statistical baseline.

Every month, we publish an accuracy report on your own data — error rates across 7, 14, and 30-day horizons, raw forecast vs AI-corrected, side by side. You see exactly how the AI performs. No magic. No marketing claims.

Deep dive in Forecasting
app.peaqplus.com/forecasting/accuracy
MAPE · 12-week rolling
3 tiers · your hotel
T0T1T2 · AI
10%8%6%4%2%W1W2W3W4W5W6W7W8W9W10W11W12target ≤ 5%
May 2026: Tier 2 MAPE 6.1% · Tier 1 baseline 8.4% · improvement +27%
Tier 0
Baseline

Quarterly statistical average. Default for all hotels — no setup required.

Tier 1
Enhanced statistical

Recency-weighted history, outlier filtering, segment-specific fallback chains, and trend correction — built per-hotel.

Tier 2
AI-corrected

Tier 1 plus a daily 60-day AI correction that reads your hotel context. Available with the AI add-on.

07
AI Report Narrative

Every report, in plain language. On click.

A floating Pulse icon on every report. Click it, and Pulse reads the data — cleaned, redacted, structured — and writes a 2-3 paragraph summary with the trends, anomalies, and what to investigate next.

Responds in the user's language. Cached for 7 days; auto-invalidated when fresh data arrives.

What it sees, what it doesn't
  • Sees: the cleaned numbers — structure intact, anomalies surfaced
  • Doesn't see: UI metadata, hotel or property names (always stripped first)
  • 23 reports ship with their own narrative recipe — one-click on every one
  • Keyboard shortcut for power users · responds in seconds
app.peaqplus.com/reports/performance
Performance · March 2026
Row 117k
Row 220k
Row 323k
Row 426k
P
AI NARRATIVE
⌘⇧J

Revenue tracking 9% above plan month-to-date, led by OTA channels (+18% YoY).

Watch: corporate segment lagging −12%, particularly in W17. Recommend investigating direct cannibalization.

Cached · expires in 6d
08
Daily Briefing

Your morning summary. In your inbox by 7 AM.

The Daily Briefing is not a dashboard. It's an 8-block written report — the kind of summary your assistant would write if you had one. Yesterday's pickup. Today's OTB. Next 7 days. Monthly trajectory. Year-over-year same-point comparison.

With Pulse, two AI-generated blocks: 1-3 anomaly warnings, plus the top 3 actions for today — ranked.

Where it lives
  • Email — every morning to your team
  • In-app — Today tab → "Daily Briefing" page
  • 5 base blocks + 2 AI blocks (Pulse add-on)
DAILY BRIEFING
Hotel Aurora · March 5, 2026
Generated 07:00 CET · 8 blocks
Yesterday pickup
42 RN
+12% vs LY
Today OTB
€684k
+8.4%
NEXT 7 DAYS
M
T
W
T
F
S
S
PULSE WARNING
Weekend pickup is 18% behind forecast. Three competitor properties added discounts overnight. Suggested action: review Fri/Sat rates by EOD.
Blocks 1–5 base · Blocks 6–7 require Pulse add-on
09 · A day in Pulse

Six conversations. Six minutes.

The hotel manager doesn't memorise reports or tool names. They have a question, and Pulse runs the right query against live data. These are the conversations from a typical week.

01
Morning routine

"What happened yesterday?" Pulse pulls yesterday's close, today's OTB, the next-7-day forecast, month progress, 7-day momentum, and the same-point YoY view — onto one screen.

5 minutes · one screen
02
Operational decision

"Which days need a marketing push?" Pulse returns the specific dates with low-occupancy risk and suggests a next step — "Draft ads for these dates".

Targeted, not generic
03
Strategic review

"How are we vs Q2 plan?" Pulse answers with pace %, the room-night gap, the revenue gap, the ADR needed to close it, and the same point last year.

Actionable, not "fine"
04
Channel analysis

"Which OTA brought the most this month?" Pulse runs the segment breakdown — share, ADR, room nights — channel by channel.

Per-OTA visibility
05
Anomaly detection

"Are there days where ADR is 25% off the average?" Pulse lists the triggered days and offers to save the question as a scheduled alert.

Promote ad-hoc to scheduled
06
Booking curve

"How did Aug 20 fill up?" Pulse builds the time-series of snapshots, showing the booking curve as it filled.

Retrospective shape
The fine print

AI you can audit.

We care more about what Pulse can't do than what it can. Every safeguard is in code, not in the prompt — because prompts can be jailbroken, and code can't.

Memory stays with your hotel

Each hotel's memory lives in its own private space. The boundary is built into the data layer, not a policy you have to trust — so memory from one hotel structurally cannot reach another.

Enforced in the data layer
Sensitive data redaction

Hotel names, property names, and other identifiers are stripped before any prompt is built. The model sees the shape of the business — never the name.

Redaction · before every prompt
$
AI usage included

AI calls don't bill back to your hotel. Usage is part of your Pulse subscription — no per-call charges, no usage-based surprises, no metered billing. The internal cost guards inside the platform are how we keep that promise on our side, even on heavy-usage days.

Flat-rate pricing · usage included
Structured output

The forecast AI can only respond through a single structured contract. There is no free-text path back — parse errors are statistically near-zero.

Schema-locked responses
Range guards

±30% deviation clamps on AI corrections. Room-count caps. Guards live in code, not in the prompt — so no clever question can bypass them.

Code-enforced · not promptable
Audit trail

Every AI call — chat, forecast, narrative, briefing — is logged with cost, outcome, and the full request and response. Reviewable on request.

Per-call audit log
10
The stack

The right AI for each job.

No single AI is best at everything. The forecast correction needs deep numerical reasoning; the chat needs to be fast; the narrative needs to fit the user's language and stay concise. We pair each Pulse capability with the model that fits the work — and swap providers when a better one ships, without changing how you use the product.

What you get
  • Language follows the user — every Pulse feature answers in the user's account language
  • Real-time chat with offline-safe fallback — the answer always arrives
  • Cost-tracked per hotel — see exactly what AI is spending on your behalf
  • Provider-swappable — model is a setting, not a rewrite
Smart Forecast Enhanced
Frontier reasoning model
Deep numerical reasoning across 60 days of context, structured output
Pulse Chat
Fast multi-turn model
Low latency, tool-calling at scale, cost-efficient per conversation
AI Report Narrative
Multilingual narrative model
Summaries in the user's language, fast turnaround on every report
Daily Briefing blocks
Structured short-form
Compact, factual paragraphs — the same shape every morning
Signal → Decision → Action → Outcome

Meet Pulse. With your own data.

In our 45–60 minute walkthrough (length depends on how deep you want to go), we run Peaqplus on our live demo environment — a simulated property with data that moves day to day — ask Pulse a few questions, and show you what an AI Revenue Manager looks like when it speaks hotel.

No setup fee. Pulse only activates after you opt in.