Mike HeapMike is an experienced Product Manager who focuses on all the “non-development” areas of My AskAI, from finance and customer success to product design, copywriting, testing and more.
Table of Contents
- What is Kustomer (and Kustomer AI)?
- How easy is it to set up Kustomer AI?
- What channels does Kustomer AI work in?
- What are the limitations of Kustomer AI?
- What knowledge sources can I train Kustomer AI on?
- What features does Kustomer AI have?
- How do I improve Kustomer AI responses?
- What resolution rate can I expect from Kustomer AI?
- What AI model does Kustomer AI use?
- What languages does Kustomer AI work in?
- How secure is Kustomer?
- Who is using Kustomer?
- How much does Kustomer cost?
- Seat-based plans
- Conversation-based plan
- AI Agent add-ons
- Does Kustomer have a free trial?
- Is Kustomer worth it?
- What are the Pros and Cons of Kustomer?
- Pros
- Cons
- FAQs
Created time
Jun 30, 2026 10:30 AM
Title length (<60)
Author
Last optimised
Ecomm?
Image
Publish date
Jul 2, 2026
Video
Slug
kustomer-complete-guide-2026
Featured
Type
Article
Ready to Publish
Ready to Publish
Kustomer says its AI Agents resolve up to 70% of inquiries. The real cost: $0.60 per engaged conversation, on an 8-seat-minimum annual contract. Here's the full picture.
Someone's asked you to "go and look at Kustomer's AI", and you've spent twenty minutes on the website without finding the one thing you actually came for: the price. The marketing says its AI Agents resolve up to 70% of customer inquiries. The pricing page says "Talk to Sales".
That gap, between the big claim and the buried detail, is the whole experience of sizing up Kustomer right now (I've watched a lot of buyers hit that exact wall). It's a real, well-funded, enterprise-grade product with a proper CRM behind it.
It's also one of the pricier ways to buy AI customer service, and the cost is annual, sales-led, and starts at an 8-seat minimum.
You're probably here for one of three reasons:
- You already run on the Kustomer CRM, and you're deciding whether to switch on its AI Agents…
- You're on Zendesk, weighing Kustomer's new standalone AI overlay against an agent that sits on the stack you already have…
- You're trying to model the real monthly cost for your ticket volume, and the "Talk to Sales" wall keeps getting in your way.
Either way, I've got you. This guide breaks down what Kustomer (and Kustomer AI) actually is, how it works, where it fits, what it really costs, what resolution rate to expect, and how to decide if it's right for you, minus the marketing gloss.
What is Kustomer (and Kustomer AI)?
TL;DR: Kustomer is an AI-native customer-service CRM built on a unified customer timeline. Its "Kustomer AI" layer adds an autonomous agent and an agent copilot, and since March 2026 it can overlay an existing Zendesk.
Kustomer is an AI-native customer-service CRM. In support-team terms, it's three things: a unified customer timeline that pulls every interaction into one view, an autonomous AI agent, and a copilot for your human agents. All of it sits on the CRM underneath.
The AI bit is branded "Kustomer AI", and it splits in two. There's AI Agents for Customers (the customer-facing agent that resolves tickets on its own) and AI Agents for Reps (a copilot that drafts replies and summaries for your team). Both run on what Kustomer calls its AI Reasoning Engine.
The naming has some history worth knowing, because the old labels still pop up in reviews. Kustomer's original AI brand was "Kustomer IQ" (KIQ), launched back in 2021, and the first chatbot, KIQ Conversations, was retired on June 30, 2025.
KIQ Customer Assist and KIQ Agent Assist then got folded into today's "AI Agents for Customers" and "AI Agents for Reps". So if you spot "KIQ" in an older review, it's the same family wearing an old name.
The company itself dates to 2015. It was founded by Brad Birnbaum (the CEO, who previously co-founded Assistly, later acquired by Salesforce) and Jeremy Suriel (the CTO), and it runs about 300 people out of New York.
On G2, one feature gets praised far more than any other: that timeline.
G2: Kustomer scores 4.5/5 from 493 reviews on G2. "Kustomer's greatest strength is its unified customer timeline, which consolidates every interaction across chat, email, and other channels into a single, chronological view… eliminating the need to switch between platforms." via a G2 reviewer.
Here's the question I'd sit with first, though. A lot of teams just want an AI agent, and they don't need a whole new CRM to get one.
If you don't need a full CRM, My AskAI puts an AI agent on the helpdesk you already run (Zendesk, Intercom, Freshdesk, Gorgias, or HubSpot), so you keep your stack instead of moving platforms.
How easy is it to set up Kustomer AI?
TL;DR: You build the agent in a no-code studio, and Kustomer markets a 15-minute Zendesk go-live. The headline automation rates take real configuration and ongoing tuning, so read 15 minutes as time-to-first-reply.
You build the agent in AI Agent Studio, a no-code visual builder, using a "Define, Ground, Equip, Gain" framework: define the agent's role and guardrails, ground it in your knowledge base, give it tools, then watch how it behaves through evaluations. You write the instructions in plain language, "just as they would with a ChatGPT prompt".
For the Zendesk overlay specifically, Kustomer markets a 15-minute go-live, "no engineering needed". In February 2026 it also shipped an AI Setup Assistant that connects to Zendesk, reads your existing config, and rebuilds it inside Kustomer (handy if you're moving over from Zendesk).
The reality is a bit softer than "15 minutes". Hitting the 40% automation rate Kustomer cites in its flagship case studies takes real work: knowledge-base tidying, instruction tuning, and ongoing evaluation.
The go-live is fast. The results come from the setup effort, and flipping a switch won't get you there. I'd read that 15-minute number as time-to-first-reply; time-to-target is a longer haul.
We set My AskAI up in about 10 minutes on your existing helpdesk, with no developer needed for the standard setup, and the agent you train moves with you if you ever change helpdesks.
What channels does Kustomer AI work in?
TL;DR: Chat, email, SMS, WhatsApp, Instagram, Facebook Messenger, web forms, and natively-built AI Voice, which Kustomer claims as a category first.
Kustomer covers the full omnichannel spread: chat, email, SMS, WhatsApp (via Twilio or MessageBird), Instagram, Facebook Messenger, and web forms. It also runs AI Voice built natively into the platform rather than bolted on through a third party. That's a genuine strength, and something most pure-play AI vendors don't offer.
Kustomer claims to be the first provider to do native conversational AI on voice without a third-party integration. I'd take the "first" as marketing language (Sierra and Decagon make their own versions of that claim), but the native-voice capability itself is real and still pretty uncommon.
My AskAI replies in whatever channels your helpdesk routes to it, from web chat and email through to the social channels your helpdesk already handles.
What are the limitations of Kustomer AI?
TL;DR: The recurring four are a steep learning curve, long-standing reporting weaknesses, about three outages a month, and resolution claims that no independent benchmark backs.
This is the section the marketing pages won't hand you, so it's the most useful one. Kustomer is capable, but four limits come up again and again.
- The learning curve is steep. On G2, the most-tagged complaints are "Learning Curve", "Complexity", and "Not Intuitive". A Gartner reviewer sums it up nicely: "The vast array of feature and customization options may overwhelm those who are not familiar with customer service platforms." The flip side of all that flexibility is that you'll want someone technical to make the most of it.
- Reporting has been a weak spot for years. It's one of the most consistent gripes across review sites: inflexible reports, numbers that shift day to day, and a wait before you can trust the data. Kustomer added a natural-language Data Explorer in late 2025 to help, but the reporting reputation is well baked in by now.
- Reliability isn't spotless. The uptime tracker IsDown has logged roughly 180 outages over five years, about three a month. For a system running live customer conversations, that's worth a second look.
- The autonomous-AI claims are thinly checked. Most of the positive AI commentary in reviews is general ("AI-powered suggestions help productivity"), with little that speaks to the autonomous agent specifically, and there's no independent benchmark of Kustomer's resolution rates anywhere. Every 40%/70% figure is vendor-supplied or quoted inside a Kustomer case study. None of that makes the product bad. I'd just prove the resolution rate on your own tickets before trusting the headline number.
One more gap for the ecommerce crowd: Kustomer integrates deeply with Shopify, Magento, and BigCommerce, but WooCommerce is nowhere to be seen.
When a My AskAI reply looks off, you can ask Echo why the agent gave that answer and which knowledge source it used, so the reasoning behind every reply is there for your team to check.
What knowledge sources can I train Kustomer AI on?
TL;DR: Native KB, public URLs, third-party systems via OpenAPI tools, Shopify, a crawler that imports a full Zendesk Help Center, and an MCP client and server.
Kustomer's AI Agents can ground their answers in your native Kustomer knowledge base, public URLs, and third-party systems via OpenAPI tools. For ecommerce there's that deep Shopify connection.
A Fall 2025 release added a web crawler that imports an entire Zendesk Help Center, handy now that Kustomer overlays Zendesk. It also comes with an MCP (Model Context Protocol) client and server, so it can both pull in external context and expose Kustomer data to assistants like Claude and ChatGPT.
Your knowledge stays in your own private workspace and isn't shared across other customers (useful context for the security review further down).
My AskAI trains on more than a help center: Google Drive, Notion, Confluence, SharePoint, Shopify and your past tickets all feed the knowledge it answers from. No help center yet? Training on historic tickets builds you a starter knowledge base from past resolved tickets.
What features does Kustomer AI have?
TL;DR: Three layers: an autonomous agent, an agent-facing copilot, and an actions and workflow layer with bidirectional AI-to-human handoff, governed by a Supervisor AI and automated evals.
Kustomer's AI splits into three layers.
- The autonomous agent (AI Agents for Customers) handles order status, refunds, reservation changes, account updates, and policy questions, and it can run multi-step interactions that would normally need a human, escalating only when the situation genuinely calls for it. It reaches into external systems through OpenAPI tools to fetch live data and complete actions.
- The copilot (AI Agents for Reps) sits with your human agents, drafting replies and summarizing long threads. Kustomer claims a 65% boost in agent efficiency from it.
- The actions and workflow layer lives on the CRM backbone: a no-code Workflow Builder, OpenAPI-spec'd tools, and (a real differentiator) bidirectional handoff, where a human can pass a conversation back to the AI to finish off after dealing with the tricky part. Governance comes from a "Supervisor AI" that reviews team responses before they reach a customer, plus automated evaluations that grade conversations for accuracy and tone. In December 2025, Kustomer expanded this into an AI Assistants family of six tools.
My AskAI runs refunds, cancellations and order updates through Tasks, and you choose per action whether the AI completes it itself or drafts it for an agent to approve.
How do I improve Kustomer AI responses?
TL;DR: Monitor, evaluate, fix, re-test, backed by automated evals, an Observability Assistant, and Data Explorer. It rewards ongoing tuning over set-and-forget.
The improvement loop is monitor, evaluate, fix, re-test. Kustomer's automated evaluations grade conversations at scale, and an Observability Assistant flags where things went sideways: bad tool calls, fuzzy instructions, missing context, or the wrong knowledge article getting pulled. The Data Explorer then surfaces patterns in plain language.
The bit I'd plan for: this isn't a set-and-forget product. The teams that hit the headline automation rates are the ones that keep tuning knowledge and instructions over weeks and months. Budget for that ongoing effort on top of the initial build.
Self-Learning drafts new knowledge articles each week by comparing the AI's replies to your agents', so the knowledge base keeps improving with less hands-on work from you.
What resolution rate can I expect from Kustomer AI?
TL;DR: Kustomer claims up to 40% to 70%, but no independent benchmark validates it, and it bills per "engaged conversation", which isn't a resolution. Treat the number as a vendor claim until you test it on your tickets.
Kustomer's public claims range from "up to 40%" of inquiries resolved (the conservative figure in its Series B announcement) to "up to 70%" on the current marketing page. Both numbers turn up without a like-for-like definition.
The named case studies give you a more grounded read (remember, these are the wins Kustomer chose to publish). Vuori reported 40% of chat conversations fully automated, and Everlane cited a 4x jump in live-service deflection, with 10% of chat handled without an agent on day one. Aplazo, a Mexican BNPL provider, reported a 40% CSAT lift and a 60% automation share in early 2026, and Cruisebound says 80% of its customers now self-serve.
Two cautions I'd flag, though. First, there's no independent benchmark; all of these are vendor-published. Second, watch the metric: Kustomer bills per "engaged conversation", which isn't the same thing as a resolution.
Across our own field dataset of AI handling rates (a first-party aggregate spanning roughly 55 vendors and 195 deployments) the median sits at 70%, but the label moves the number more than the capability does: "resolution" runs about 12% higher than "automation" simply because each vendor counts a different event. (These figures are aggregate and directional; every vendor defines its own metric, so read them as context.) So when you see "up to 70%", my advice is to ask what it's counting before you compare it to anything.
My AskAI averages a 72% resolution rate across its customer base, and because you pay per ticket the cost per resolved ticket falls as that rate climbs.
What AI model does Kustomer AI use?
TL;DR: OpenAI, Anthropic, and AWS Bedrock, without per-feature disclosure. Generative features default to OpenAI's GPT-4o-mini.
Kustomer names its model providers but not which model powers which feature. Its AI compliance page lists OpenAI, Anthropic, and AWS Bedrock, and says outright that several LLMs might run at once. As of mid-2024, the generative features default to OpenAI's GPT-4o-mini, an upgrade from GPT-3.5-Turbo. (Fun fact: despite the years under Meta's ownership, there's no Llama anywhere in the stack.)
What languages does Kustomer AI work in?
TL;DR: 53 languages for outbound translation via Amazon Translate. Native conversational coverage is narrower, around 25 to 35.
Kustomer supports 53 languages for outbound translation, powered by Amazon Translate. That figure is about translation rather than native conversational coverage, though, and third-party estimates put the AI Agent's native conversational range closer to 25 to 35.
If multilingual support is core to your operation, I'd confirm the exact languages your customers use during the trial.
My AskAI handles 95 languages, auto-detected per message, with live translation for agents who can't read the customer's language.
How secure is Kustomer?
TL;DR: SOC 2 Type II, ISO 27001, GDPR, CCPA, and HIPAA with a BAA (a $25/user/month add-on). AWS-hosted, with US and EU data regions.
Kustomer holds the certifications an enterprise security review will ask about: SOC 2 Type II and ISO 27001, plus GDPR and CCPA compliance. It supports HIPAA with a Business Associate Agreement, sold as a $25/user/month add-on. The platform runs on AWS with US and EU data regions (the EU one has been in Dublin since 2019).
On AI data handling specifically: customer data isn't used to train global models, OpenAI works under a zero-retention policy, and AI-generated chat replies carry a "Generated by AI" label. One housekeeping note for reviewers: Kustomer's security page still lists the EU-US Privacy Shield framework, which Schrems II invalidated, so confirm the current transfer mechanism directly.
My AskAI is SOC 2 Type II certified and GDPR-compliant, and your team can use Echo to trace why the AI gave any answer when a security or CX review asks.
Who is using Kustomer?
TL;DR: More than 600 brands, weighted to DTC retail, travel, and consumer services: Vuori, Everlane, SKIMS, Hopper, Priceline, Ring, and more.
Kustomer reports more than 600 brands, weighted heavily toward direct-to-consumer retail, travel, and consumer services. Named customers include Vuori, Everlane, SKIMS, Reformation, Glossier, and Rent the Runway in apparel; Sweetgreen, Daily Harvest, and Terra Kaffe in food and drink; HexClad, Caraway, and Jerome's Furniture in home goods; Hopper, Priceline, Turo, and Cruisebound in travel; plus Ring, Away, and (over in Latin America) Aplazo, Rappi, and Glovo.
The pattern is clear. If you're a mid-market or enterprise DTC, retail, or travel brand, you'll find plenty of peers on the platform. If you're a smaller SaaS or service team, you'll find fewer.
How much does Kustomer cost?
TL;DR: Seat-based plans run $89 to $139 per user per month on an 8-seat minimum, and the autonomous AI is a $0.60-per-engaged-conversation add-on. There's no free trial.
This is the question the website dodges, so I've pieced the full picture together from third-party reporting and Kustomer's own buried pricing-detail page. There are two parallel models, both sales-led, with no self-service signup.
Seat-based plans
This is the dominant model. Enterprise runs $89 per user per month and Ultimate $139 per user per month, both billed annually with an 8-seat minimum.
Even a 5-agent team pays for all 8 seats, so your real minimum is $8,544 a year on Enterprise or $13,344 on Ultimate. There's no monthly billing and no free tier.
Conversation-based plan
Launched in late 2024, this charges $0.35 per conversation (Enterprise) or $0.50 (Ultimate), with unlimited users and AI included. A "conversation" is billed whatever the outcome.
AI Agent add-ons
On the seat-based plans, the autonomous agent is a separate line item. Per Kustomer's Comprehensive Pricing Details page:
Add-on | Price |
AI Agents for Customers (autonomous) | $0.60 per engaged conversation |
AI Agents for Reps (copilot) | $40 / user / month |
HIPAA compliance | $25 / user / month |
Kustomer Voice | from $0.02 / minute |
The phrase I'd underline is "engaged conversation". Kustomer charges $0.60 any time the AI responds, whether or not it resolves a thing.
If the AI takes one look and bumps the ticket to a human, you still pay. That's a real gap from per-resolution pricing. It works against you.
Put it together, and a 10-agent team on Enterprise pays roughly $28,000 a year before any AI-Agents-for-Customers usage (and that's before the autonomous agent handles a single ticket). Here's the seat-and-copilot floor:
Assumption | Value |
Plan | Enterprise ($89/user/mo) |
Agents | 10 |
Seat cost | $890 / month |
AI Agents for Reps (copilot) | $40 × 10 = $400 / month |
Autonomous AI | $0.60 × engaged conversations (extra) |
Annual seat + copilot floor | ~$15,480 / year |
The pricing complaints write themselves. Reviewers call it "quite pricey", aimed at companies "that can afford 8 or more seats", and point out there's "no way to try Kustomer on a monthly basis".
We charge about $0.10 per ticket at My AskAI, flat, whether or not the ticket resolves, with the AI bundled into the plan. At 10,000 tickets a month that's roughly $1,299/month on our Scale plan, and a per-ticket bill stays flat as the AI improves while per-conversation and per-resolution bills climb.
Does Kustomer have a free trial?
TL;DR: No. No free trial, no free tier, annual billing, and an 8-seat minimum, which is the highest commitment bar in the category.
No. Kustomer offers no free trial and no free tier, bills annually, and holds you to the 8-seat minimum, which is the highest commitment bar in the category. Capterra confirms it "does not currently offer a free trial or free version". You evaluate Kustomer through a sales process and a paid annual contract.
If testing before you buy matters, My AskAI runs a 30-day free trial with every feature unlocked, unlimited tickets, and no card, so you can prove the resolution rate on your own tickets before paying a penny.
Is Kustomer worth it?
TL;DR: Worth it for mid-market and enterprise DTC, retail, and travel brands that want a full CRM and can clear the 8-seat commitment. A poor fit for smaller teams, or anyone who just needs AI on the helpdesk they already run.
It depends entirely on who you are. Here's how I'd split it.
Choose Kustomer if…
- You're a mid-market or enterprise DTC, retail, or travel brand.
- You genuinely want a unified CRM, native voice, and omnichannel under one roof.
- You can clear the 8-seat, annual commitment.
- You've got the technical resource to configure and tune it.
For that buyer, the depth of the customer timeline is hard to beat.
Don't choose Kustomer if…
- You're a smaller team without 8 agents to fill the seat minimum.
- You want transparent self-serve pricing you can model yourself.
- You simply need an AI layer on the helpdesk you already run.
The 8-seat floor, the annual contract, and the sales-led pricing all push it out of reach for SMB and smaller mid-market teams.
That last group is worth dwelling on, because Kustomer's March 2026 strategy speaks right to it. Kustomer launched its AI as a standalone product that overlays an existing Zendesk without a migration, a clear pivot from "replace your helpdesk" to "be the AI layer on top of it". It's a smart move, I think, that clears Kustomer's biggest old barrier, but it also drops Kustomer into a crowded field of AI agents that already sit on your existing stack.
And that's the real call for a lot of you. If what you want is AI on the helpdesk you already chose (Zendesk, Intercom, Freshdesk, Gorgias, or HubSpot) you don't have to adopt a second CRM's AI layer to get it.
An agent that lives inside your helpdesk (the bet we've made at My AskAI) keeps your stack, your macros, your routing, and your agents exactly where they are. And the work you put into training it travels with you if you ever switch helpdesks, so you're never locked into one vendor's platform.
For teams that picked a best-of-breed helpdesk on its merits, that's usually the better fit.
What are the Pros and Cons of Kustomer?
Pros
- A deep CRM backbone and unified customer timeline. The most-praised feature in Kustomer's reviews, and structurally hard for pure-play AI vendors to copy.
- Genuinely native AI Voice and bidirectional AI-to-human handoff. Real product wins over deflection-only chatbots (see Features).
- Buyer-friendly conversation-based pricing and the new Zendesk overlay. AI is included in the conversation rate, and the March 2026 overlay clears the old rip-and-replace barrier (see Cost and Worth it).
Cons
- High entry cost and pricing opacity. 8-seat minimum, annual-only billing, no trial, and seat prices hidden behind "Talk to Sales" (see Cost).
- A steep learning curve and configuration burden. Reviewers say making the most of its flexibility needs technical help (see Limitations).
- Persistent reporting weaknesses and roughly three outages a month. Both are long-running review themes (see Limitations).
- Thinly validated autonomous-AI claims. No independent benchmark backs the 40 to 70% resolution figures (see Resolution rate and Limitations).
Kustomer
- Brand: Kustomer (independent; Meta is a minority shareholder)
- Rating: 7/10
- In a sentence: a credible, well-funded, enterprise-grade AI-CRM that's strong for mid-market DTC retail and travel, but it carries the category's heaviest baggage (complexity, pricing opacity, and a high entry cost) that a leaner, more transparent option can exploit.
FAQs
How much does Kustomer AI cost?
Kustomer's autonomous AI Agent runs $0.60 per engaged conversation, charged any time the AI responds, whether or not it resolves the issue. That sits on top of seat-based plans at $89 (Enterprise) or $139 (Ultimate) per user per month, billed annually with an 8-seat minimum, so the real entry point is around $8,544 a year before any AI usage. There's no free trial.
What does the company Kustomer do?
Kustomer is an AI-native customer-service CRM. It pairs a unified customer timeline (every interaction across channels in one chronological view) with autonomous AI Agents that resolve tickets and a copilot that helps your human agents, all on a CRM backbone built for retail, travel, and consumer brands.
Is Kustomer a good company?
Kustomer is a well-regarded mid-market and enterprise platform, scoring 4.5/5 across nearly 500 G2 reviews, with its unified timeline the standout. The common knocks are a steep learning curve, weak reporting, and high, opaque pricing, so it's strong for the right buyer (DTC retail and travel at scale) and a poor fit for smaller teams.
Is Kustomer part of Meta?
Not any more. Meta acquired Kustomer for around $1 billion in 2022, then spun it out in May 2023 at roughly a $250 million valuation. Kustomer is now an independent, venture-backed company (Meta keeps a minority stake) and raised a $30 million Series B led by Norwest in August 2025.
What is Kustomer used for?
Teams use Kustomer to run customer support across chat, email, voice, SMS, and social from a single CRM-backed timeline, and to automate a chunk of those conversations with AI Agents that can answer questions and take actions like refunds and order updates.
What is Kustomer IQ?
Kustomer IQ (KIQ) was Kustomer's original AI brand, launched in 2021. Its products, KIQ Customer Assist and KIQ Agent Assist, have since been rebranded into today's "AI Agents for Customers" and "AI Agents for Reps". So if you see "KIQ" in older reviews, it's the same AI family under its previous name.
Does Kustomer have a free trial?
No. Kustomer has no free trial and no free tier. You evaluate it through a sales process and a paid annual contract with an 8-seat minimum.
What channels does Kustomer AI support?
Chat, email, SMS, WhatsApp, Instagram, Facebook Messenger, web forms, and natively-built AI Voice. That native voice (without a third-party integration) is one of Kustomer's genuine differentiators.
How many languages does Kustomer support?
Kustomer supports 53 languages for outbound translation via Amazon Translate. Native conversational AI coverage is narrower; third-party estimates put it around 25 to 35 languages, so I'd confirm your specific languages during evaluation.
What AI models does Kustomer use?
Kustomer uses several providers (OpenAI, Anthropic, and AWS Bedrock) without saying which model powers which feature. As of mid-2024, its generative features default to OpenAI's GPT-4o-mini.
Is Kustomer HIPAA compliant?
Kustomer supports HIPAA compliance and will sign a Business Associate Agreement, offered as a $25/user/month add-on. It also holds SOC 2 Type II and ISO 27001 certifications and complies with GDPR and CCPA.
What are the best Kustomer alternatives?
The strongest alternatives depend on what you're replacing. If you want the CRM depth, enterprise AI platforms like Decagon, Ada, and Sierra compete on the Zendesk installed base. If you'd rather add AI to the helpdesk you already run (Zendesk, Intercom, Freshdesk, Gorgias, or HubSpot) a per-ticket agent like My AskAI skips the platform switch entirely.
