For mid-market customer-ops teams running 5+ disconnected tools

Support that fixes the product — not just the ticket.One AI brain across support, retention, engagement and voice — the only one that ships the fix, not just the reply.

It knows every customer — their history, ARR, and your roadmap — and acts across the whole relationship: it answers them, keeps them, reaches them, and calls them — then turns every edit and complaint into a shipped fix, eval-gated and auto-rolled-back. Self-improving every six hours, behind your safety gates.

Support · Retention · Engagement · Voice — across email, web chat, WhatsApp, push, SMS, App Store & more

Free pilot on your own ticketsBook a 30-min demo

Book a walkthrough — set up with us, live in a week. Or see every product →

app.pilotpm.ai/inbox
Account locked after update
sara.lin@acme.io · 2m ago
Reply
Hi, I can't log into the app on my phone. It says "Account locked" after the latest update. Can you unlock it?
✦AI Draft· awaiting approval0.78
Hi Sara — sorry about that. I've unlocked your account; last week's release introduced a session bug that's being patched in tomorrow's update. Try logging in again now…

Their AI answers tickets.Ours improves the company that answers tickets.

Most AI support tools stop when the ticket closes. PilotPM turns what your team fixes by hand into fixes that ship — see it side by side vsIntercom,Zendesk,Freshdesk,Productboard andGainsight.

Works with your stack

SlackHubSpotJiraLinearGitHubWhatsAppInstagramMessengerLINEApp StoreGoogle PlaySentrySnowflakeShopifyStripeAirtableGoogle WorkspaceZaloAPI-push (custom)
“PilotPM is the single AI brain behind our entire customer surface — email, App Store + Play reviews, social, and HubSpot-synced CSM and sales out of Slack. In two weeks, the share of drafts my team sends without changing a single word doubled — from about 20% to over 40% — because every edit we make teaches it. And when something genuinely needs code, it reaches engineering as a clean, reproducible ticket — often with a draft PR and full context.”
ELSA Speak
Akshaya Aradhya, CTO
ELSA Speak
2× in two weeksdrafts sent word-for-word: ~20% → 40%+
89% resolvedwhere the AI answers alone — a takeover counts as a failure
$300K+
The stack it replaces — help desk to call center
6→1
Help desk · CS · engagement · voice · roadmap · VoC — one brain
30 min
Signup to first AI reply
One brain, every job

One customer brain. The whole relationship.

Every other vendor sells you one box — a help desk, a success platform, a marketing tool — each with its own copy of your customer. PilotPM is one AI brain that knows the whole customer and acts across all of them. See every product →

Answer

Support

AI answers every inbound message with full customer context, resolves what it can end-to-end, and routes the rest — across email, web chat, WhatsApp, App Store and every channel your customers use.

Included in every plan
Keep & grow

Retention

Your customer-success platform, made agentic. Health, renewals, expansion and NPS — with a goal-based agent that picks the next best action for each account, instead of another dashboard of rules to maintain.

From $499/mo
Reach

Engage

Lifecycle messaging at consumer scale — email and in-app — targeted on what each customer actually did, on the same profile your support inbox reads. Surveys are included. Consent, quiet hours and suppression are built in.

From $2,500/mo
Gather

Community

A place where your customers answer each other, moderated by the same AI that drafts your support replies. An accepted answer joins your knowledge base. Billed on members who showed up this month, not on everyone who ever registered.

From $250/mo
Talk

Voice

An AI-native softphone built into the inbox. Every call is transcribed, summarized and turned into a ticket with the next step already drafted — and the brain can place a call as one of its actions.

$50/seat + $0.06/min
Acquire

Ads

Reads your Google Ads and Apple Search Ads accounts nightly, verifies every platform-reported conversion against the leads and signups that actually arrived in your own data, and turns what it finds into recommendations you approve in one click — applied automatically under hard budget rails, with a full audit trail and one-click revert.

From $199/mo
Guide

Genie

A white-label companion your end users talk to inside your app — it knows their profile and progress, answers how-to and how-am-I-doing questions in their own language, voice included, and hands off to your support team with identity attached when a human is needed.

From $500/mo
How it works · Customer support automation

AI resolves it. Your team owns what matters.

Every inbound message gets a reply drafted in seconds with full context — the customer's history, their ARR, your roadmap, last quarter's commitments. High confidence: it sends. Risky: it queues. Always with citations.

How a reply gets drafted — five steps, four seconds

Consolidates support inbox, KB, and ticketing · $80–160K/yr
01

Inbound lands

Email, chat, WhatsApp, Instagram, App Store, Slack & more. PII encrypted at rest.

→
02

Context retrieved

KB, customer history, ARR, last quarter's commitments, roadmap — one prompt.

→
03

AI drafts

A cited reply with a self-rated confidence score, in the customer's language.

→
04

Confidence gate

Per-channel floors (App Store 0.95, email 0.85). Risky tokens always go to a human.

→
05

Send or queue

High confidence → auto-sends in secondsLow confidence → queued, draft already written
A branded public help site at /help/yourcompany — the same AI on every page — is included.
  • Import your docs, point your DNS — customers self-serve before they email.
  • AI-only chats are free; only escalations count against your cap.
  • Any language the customer writes — auto-detected, tone-matched.
  • Every approved reply grows your KB — the AI cites your team's answers, not a generic FAQ.
Use case 3 · Rules in plain English

You set the rules. No prompt engineering.

Reply rules let anyone — CS lead, Success manager, PM — write natural-language guidance the AI follows on every draft. Auto-pilot decides where AI sends on its own and where it queues for human review.

Reply rules

Write what the AI should do.

Each rule is a sentence. Scoped to a channel, lens, customer tag, or workspace-wide. The drafter sees the last 20 messages of the thread, so rules can be conditional on history.

Example rule
“If the customer has already asked this question in this thread and we pointed them at an article, offer to connect them with a person and set escalation to 'escalate'.”
Auto-pilot

AI sends. You set the bar.

Off by default. Flip it on, set your confidence threshold, and the AI sends without human review when it's above the bar — except where it shouldn't.

  • Hard blocks regardless of confidence: refunds, account cancellations, passwords, lock/unlock, money, percentages, contracts, "talk to a human". One opt-in exception: purely instructional “how do I cancel?” steps can auto-send if you enable it — anything touching the account, an amount, or an offer still always goes to a person.
  • Per-channel floors above your setting: App Store + Instagram require 0.95
  • Public channels need 100+ human-approved drafts before auto-send unlocks (calibration gate)
Trust & security

Your data. Your tenant. Your control.

Tenant isolation on every query

Every read and write is scoped to a single workspace — your data is only ever returned within that scope.

Encrypted everywhere

AES-256-GCM at rest. TLS 1.3 in transit. HMAC-fingerprinted user IDs, never raw alongside conversation data.

Audit trail per conversation

Every AI action, every approval, every signal traced. Full audit log exportable for security reviews on Business+.

Right to be forgotten

One-click deletion cascades through derived data, embeddings, KB entries. GDPR + CCPA compliant. DPA before contract.

Your data trains your AI. Not anyone else's.

Tenant isolation goes all the way down — model calibration, embeddings, save-offer history, draft acceptance signals all stay scoped to your workspace.

Continuous compliance

Agent-driven control checks running 24/7 across data handling, access, and DPA terms — not a once-a-year stamp.

Continuous controls monitoringGDPR-alignedCCPA-alignedHIPAA-ready

Pricing

Pick your tier by capability. Pay for AI work as you use it.

One credit pool. Use it however your team needs.

Tiers buy capability — seats, lenses, advanced features. Credits buy AI work. Every action — a draft, an Ask Q&A, a roadmap suggestion — pulls from one unified pool. Allocate as you go. No stranded credits in the wrong bucket.

StarterSmall CS team
RecommendedGrowthMid-market ops
BusinessAll 5 lenses
Price
$399/mo
$2,499/mo
$7,499/mo
Credits / mo
1,000
15,000
80,000
Seats
5
20
50
Lenses
3 · + Support, Success
4 · + Product
5 · + Sales
Auto-pilot
✓
✓
✓
CS ops — routing · SLA · scorecards
—
✓
✓
Weekly AI coaching
—
—
✓
SSO
—
✓
✓
Audit log
90-day
90-day
1-year
Get started
Get started
Get started
Book a demo
Enterprise — regulated · custom contractUnlimited seats · 7-year audit retention · committed-use credit discounts · Compliance lens + custom DPA · regional / on-prem hosting · named engineer.
Talk to us
Starter
Solo founder · small CS team
$399/ mo
No setup fee
  • 5 seats
  • 3 lenses (CEO, Support, Success)
  • 1,000 credits / month
  • Auto-pilot included
  • Branded help site + outbound email
  • 90-day audit log
Get started →
Recommended
Growth
Mid-market team led by ops
$2,499/ mo
No setup fee
  • 20 seats
  • 4 lenses (CEO, Support, Success, Product)
  • 15,000 credits / month
  • Auto-assignment, SLA tracking, agent scorecards
  • SSO + 90-day audit log
Get started →
Business
Mid-market · all 5 lenses
$7,499/ mo
No setup fee
  • 50 seats
  • All 5 lenses (+ Sales)
  • 80,000 credits / month
  • B2B routing (escalate by ARR)
  • Weekly AI coaching reviews
  • Full audit-log export, dedicated CSM + SLA
Book a demo
Enterprise
Regulated · custom contract
Custom
Sized to your scale.
  • Unlimited seats + 7-year audit retention
  • Committed-use credit discounts
  • Compliance lens + custom DPA
  • Regional / on-prem hosting
  • Named engineer, quarterly founder reviews
Talk to us

Every plan is set up with us — you are live in about a week.Business runs a 30-day ROI pilot: we measure the savings together, stay or walk with no penalty.

Comparing vendors? See how PilotPM stacks up as anIntercom alternative,Freshdesk alternative, orZendesk alternative.

What you'll see in week one

Connect on Monday. By Friday the substrate's working for you.

Day 1Connect

Hook up the channels

Slack, email, HubSpot, App Store. 30 minutes of OAuth + recipient aliases. AI starts drafting replies with your KB and your last-30-days of customer voice.

Day 3Themes

First clusters appear

In Voice of Customer. Every channel's signal collapsed by meaning, ranked by ARR coverage. You see what your customers are actually asking for, in their words.

Day 7Roadmap

ARR-at-risk surfaces

Every active initiative shows ARR-at-risk drawn from real customer voice. The proposed-initiatives queue surfaces work the substrate thinks should be on your roadmap but isn't yet.

Day 30Calibration

Auto-send unlocks

On the safe categories. AI replies cite your team's actual answers, not a generic FAQ. The "Today" brief reads like the workspace knows your team. Because by then it does.

Get started

30 minutes with us, you'll know if we're it.

01

Book a demo

30 min, against your real inbound. No slides — we show the dashboard on your actual tools, and show what we'd have resolved last week.

Book →
02

Get started

A walkthrough on your own data before any commitment — we set it up with you.

Start →
03

Run a 30-day pilot

White-glove onboarding on Business tier. We measure ROI together. Stay or walk, no penalty.

Talk to us →