Customer Support, Unified.
Two tiers, still no per-agent seat tax. **Standard $49/mo** is a complete modern helpdesk — email + portal ticketing, live chat, SLA policies, knowledge base, CSAT, macros, automations, and skills-based routing. **Pro $99/mo** turns it into a full omnichannel support operation: every channel your customers actually use — SMS, WhatsApp, Instagram, Facebook Messenger, X, and phone calls — lands in the same queue as email, alongside an AI resolution agent that takes real actions, conversation QA, proactive outreach, and live translation. Unlimited agents on both, flat per-workspace, all connected to your CRM and Commerce data.
Replaces Your Disconnected Support Stack
The AI-Powered Agent Workspace
Give your support team superpowers with a deeply integrated, lightning-fast inbox.
AI Resolution Agent That Takes Actions — Pro
The flagship Pro capability, and it does more than write replies. A KB-grounded agent resolves tickets end-to-end without an agent in the loop: it reads the customer's mood first — genuinely upset customers always go to a human, never a bot — searches your knowledge base by meaning, and when it's confident, sends an empathetic, fully-formed reply that closes the ticket. Below that confidence bar it routes to a human instead of guessing.
AI Copilot for Agents — Pro
Built directly into the ticket sidebar. Real-time conversation summaries, suggested replies and draft responses grounded in your own knowledge base, similar-ticket search, and sentiment scoring so agents instantly know how urgent and how frustrated a customer is. Flags genuinely complex, multi-system issues for your senior agents instead of routing them like everything else. Every AI suggestion comes with a confidence score, so supervisors can audit them at a glance. Composes with your AeionClaw AI assistant for full cross-module context. Gracefully degrades when AI is unavailable — never a hard dependency.
AI Ticket Categorization
Standard, on by default. Automatically reads every new ticket's subject and description, categorizes it (billing, technical, account, feature request), and sets its priority — routing urgent issues straight to the right specialized queue, no human triage required.
Self-Populating Help Center
When an agent marks a ticket "Resolved", the AI summarizes the entire thread. If it determines the problem/solution pair is novel, it autonomously drafts a step-by-step Markdown article and submits it to the Knowledge Base as a Draft, so your Help Center grows itself.
Smart Reply Guard
AI intent classifier intercepts "thank you!" messages, out-of-office auto-replies, and duplicate confirmations. Keeps closed tickets closed. Protects SLA metrics from false re-opens. Only genuine new issues trigger new tickets.
Customer 360 View
Because Helpdesk shares the Aeion OS database, agents instantly see the customer's entire history. CRM Lifetime Value, active Commerce subscriptions, order status, and past support tickets—all in one panel. No tab switching.
Skills-Based Routing
Five strategies to choose from: round-robin, skill-based matching (by proficiency level), load-balanced (lowest current workload), a blend of skill and load, or fully manual. Each ticket's requirements are matched against agent skills and alternative agents are surfaced with reasons, so leads can see why a ticket went where it did. Category overrides — like always sending billing questions to the billing team — take priority over the default strategy.
SLA Policies & Business Hours
Business hours calculation (9 AM–5 PM, Mon–Fri), holiday exclusion with 1-hour cache, pause/resume timers, and immutable breach history. Tracks response and resolution targets. Automatic escalations for at-risk and breached tickets. Default targets: Critical (1hr resolution), High (4hr), Medium (24hr), Low (48hr).
Omnichannel Live Chat
Customizable widget on your website. Real-time engagement, seamless transition to async ticketing if customer leaves. Chat transcripts convert to formal tickets automatically.
Co-Browse & Screen Share — Pro
Start a full co-browse / remote-control support session directly from any ticket. Agents share screen and audio; customers join with a single click, no install required. Every session is recorded for compliance, and an AI diagnostic assistant can optionally join to help the agent spot the issue faster.
Side Conversations — Pro
Loop in engineering, billing, or any other team on a ticket via private side threads — cross-team collaboration that stays attached to the ticket without leaking into the customer conversation.
Incident & Problem Management — Pro
Link many tickets to a single underlying incident or problem record, communicate once to everyone affected, and track root-cause resolution — ITIL-style incident and problem workflows built in.
Advanced & Scheduled Reporting — Pro
Search analytics, custom report builder, and scheduled report delivery. Includes predictive SLA scoring — historical ticket data predicts which tickets are at risk of breaching before they actually breach, driving proactive reassignment and escalation.
Every Channel Your Customers Actually Use — Pro
One queue, one SLA clock, one customer history. Not five tools pretending to be one.
Social, SMS and WhatsApp Become Tickets
An Instagram DM, a Facebook Messenger message, an X DM, a WhatsApp thread or an SMS arrives and becomes a real ticket — with SLA timers, routing, macros and reporting, exactly like email. Replies from the agent go back out on the original thread, so the customer keeps talking where they started and never learns your internal tooling exists. Each ticket is badged with the channel it came in on, and the platform name for social, so an agent knows they're writing an Instagram DM and not an email before they start typing.
Delivery You Can Actually Verify
Most helpdesks tell you a reply was "sent" the moment they hand it to a provider. Ours tells you what happened to it. Every outbound reply carries a delivery receipt: sending, delivered, or not delivered with the reason — the number was unreachable, the channel can't carry outbound replies, outbound email isn't switched on yet. If the transport never confirms, it says the status is unknown rather than showing a checkmark it can't stand behind. You find out about a broken channel from the ticket, not from an angry customer three days later.
Phone Calls, Inside the Ticket
A browser softphone lives in the ticket itself — no desk phone, no separate dialer app. Click a customer's number to call them; an inbound call screen-pops their ticket and history before you say hello. Live captions run during the call, and the recording, transcript and AI summary attach to the ticket when you hang up, so the next agent reads what happened instead of asking the customer to repeat it.
Filter and Report by Channel
Filter the inbox to just WhatsApp, just social, just voice — stacked on top of your normal views. Per-channel SLA reporting answers the question every support lead eventually asks: are we actually faster on WhatsApp than on email, and where are we breaching? Set different SLA targets per channel, because a WhatsApp message and an email do not carry the same expectation of speed.
Speak Your Customer's Language
Inbound messages are detected and translated for the agent; outbound replies are translated back into the customer's language. Agents work entirely in their own language while customers read theirs. The original text is always preserved alongside the translation — nothing is overwritten — and when a translation genuinely isn't available, the system says so and sends the original rather than quietly shipping a machine-mangled reply.
Get Ahead of the Ticket — Pro
The cheapest ticket is the one nobody had to open.
Proactive Messaging
Reach customers before they contact you — a survey after a resolution, an apology before they notice a missed SLA, a heads-up during an incident. Send by email, SMS or WhatsApp, or as an in-app banner, modal, tooltip or inline note in your portal and chat widget. Fire on an event, on a recurring schedule, or as a one-off broadcast to a segment.
Guardrails That Stop You Annoying People
Frequency caps per person, quiet hours in the customer's timezone, a de-duplication window so nobody gets the same message twice, and automatic respect for anyone who has opted out. The authoring screen states the rules back to you in a sentence — "at most 1 message per person every 7 days, never between 21:00 and 08:00, skipping anyone who has opted out" — so you can check the policy at a glance instead of decoding six settings.
Know Who You're About to Reach
Before a broadcast goes out, see exactly how many people match your filter and a sample of who they are. If part of your filter couldn't be applied, it says so and warns you the audience is wider than you asked for — the difference between a message to forty VIPs and a message to your entire customer base.
Voice of Customer
Contact-reason topics are discovered automatically from what customers actually write, not from a taxonomy someone maintained once in 2023. Volume trends, statistical spike detection and a forecast per topic show you a problem forming while it's still small — and the tickets behind every topic are one click away.
Conversation Quality, Not Spot Checks
Sampling rules pull conversations for review automatically — random percentages, or targeted at low CSAT, escalations, reopens and VIP accounts. Score against a weighted scorecard you define, with critical dimensions that fail a review outright. AI drafts the scores; a human accepts or edits them. Calibration sessions let several reviewers score the same conversation to check they agree, agents can dispute a score, and per-agent coaching trends turn QA into development rather than surveillance.
Deflect Tickets with Self-Service
Empower customers to find answers before they ever click "Submit."
Semantic Knowledge Base
Beautiful, customizable self-service portal. Search understands what customers actually mean, not just the exact words they typed. Customers ask in natural language; AI surfaces the relevant articles.
Proactive KB Suggestions
When a customer starts typing in the support portal, AI instantly surfaces relevant articles inline. Deflects up to 40% of tier-1 tickets before submission.
Article Version Control
Every edit to a help article is versioned automatically, with a full change history and one-click restore to any prior version — nothing is ever lost. Time-travel further back with Aeion Aegis: roll back your entire knowledge base, or a single article, to exactly how it looked at any point in the past.
Knowledge Discovery
AI analyzes resolved tickets to suggest new KB articles, spotting patterns in customer questions that don't have a good answer yet, and promoting successful agent responses into articles automatically.
Incident Detection
AI spots systemic issues from ticket patterns — when multiple customers report the same thing independently, it raises an incident alert, catching the wave before it turns into a flood of tickets.
Public Support Portals & Live Widgets
Deliver consumer-grade support experiences directly on your public website.
Floating Chat Widget
A lightweight Live Chat bubble that drops onto your website, signed in with the same account your customers already use.
Customer Ticket Dashboard
A dedicated member area where customers can securely track their issues and browse your public knowledge base.
Standard $49/mo · Pro $99/mo — Unlimited Agents, Flat
Semantic AI Search
Tool-Using AI Resolution Agent (Pro)
Human Approval Gate on Agent Actions (Pro)
AI Copilot for Agents (Pro)
Omnichannel: SMS · WhatsApp · Social · Voice (Pro)
Verified Delivery Receipts (Pro)
In-Ticket Browser Softphone (Pro)
Conversation QA & Coaching (Pro)
Proactive Messaging with Guardrails (Pro)
Voice-of-Customer Analytics (Pro)
Inbound + Outbound Translation (Pro)
Smart Reply Guard
Predictive SLA Scoring (Pro)
Co-Browse & Screen Share (Pro)
Immutable SLA History
SOC 2 Type II ready (audit motion in flight)
Built-in AeionClaw support-agent skill
Backed by Aeion Aegis (backup + PITR + time-travel)
Real-time collaboration via Aeion Realtime
Native data import via Aeion Singularity
AI features powered by Aeion Neural Bus
Frequently Asked Questions
Every inbound ticket is scored for urgency and customer frustration, classified by category (billing, technical, account, feature request), matched to the best available agent by skill and workload, and given a drafted first reply for human review. Agents see the AI's draft pre-filled in the reply box — one-click send, or edit first.
Yes — choose from round-robin, skill-based matching, load-balanced, a blend of skill and load, or manual assignment, per queue. Agents declare their skills ("refunds", "technical", "enterprise accounts"); routing scores each agent's match and current load and picks the best fit.
Priority-based SLAs (Critical 1hr, High 4hr, Medium 24hr, Low 48hr, by default — fully configurable). The timer starts on ticket creation, pauses while you're waiting on the customer, and resumes on their reply. Breaches surface immediately in the agent dashboard and can trigger your own escalation or notification workflows.
Aeion Helpdesk ships its own live-chat widget — same backend, same agent UI as your tickets. If you need to keep an existing tool (Intercom, Drift, Zendesk Chat) during a transition, connect it via webhook so tickets flow into Helpdesk while the customer-facing widget stays put. Most teams fully consolidate within 60–90 days.
Knowledge base auto-builds from resolved tickets (with AI summarization), augmented by manually-written articles. Customers search the KB before opening tickets; deflection rate typically 30–50%. Submitting a ticket from the KB attaches the article history so agents see what the customer tried before asking.
Yes — the Forum module provides community discussion threads, voting, badges, and karma. Cross-link with Helpdesk so unresolved forum questions can convert to tickets if they go unanswered for N hours. Resolved community answers can be promoted to the KB.
Email, SMS, WhatsApp, Instagram / Facebook Messenger / X DMs, and the customer portal — reply once in the ticket and it goes back out on the thread the customer started. Live chat and voice tickets are the two exceptions: they come in as tickets with full transcripts and SLA tracking, but a text reply doesn't push back into a live chat session or a phone call, so the composer tells you that up front instead of accepting a reply that goes nowhere. Chat sessions are answered in the chat console; voice tickets are followed up by calling back with the in-ticket softphone.
It confirms what the transport reported, and it distinguishes the steps most tools blur together. **Sent** means a provider accepted the message. **Delivered** means the transport confirmed it arrived. **Not delivered** comes with the reason. If nothing ever confirms, it says the status is unknown rather than showing a tick. Notably, if outbound email isn't configured yet, a reply says so instead of appearing to have been emailed — which is the failure mode that silently costs teams customers.
It takes actions. You define the tools it can use — an order lookup, a subscription check, one of your own webhooks, or any connected system — and the agent decides when to call them across a multi-turn conversation. Anything with side effects is parked for human approval by default, showing the exact arguments, which you can edit before approving or reject outright. You can loosen that per message once you trust a particular tool.
No. Sampling rules pull conversations automatically: a random percentage of resolved tickets, and/or targeted conditions like low CSAT, escalated, reopened, or VIP customers. Each rule has a daily quota so a bad day doesn't flood the review queue, and reviewers can be assigned round-robin. AI drafts per-dimension scores against your scorecard; a human accepts or edits them before publishing, and agents can dispute a published score.
That's what the guardrails are for, and they're on by default rather than opt-in. Frequency caps limit how often any one person can be messaged, quiet hours stop overnight sends, a de-duplication window prevents repeats of the same message, and anyone who has opted out is skipped automatically. Before a broadcast you see the recipient count and a sample — and if part of your filter couldn't be applied, you're warned that the audience is wider than you intended before anything sends.
Inbound customer messages are translated for the agent and outbound replies are translated for the customer, using your configured translation provider with a reusable translation memory so repeated phrasing stays consistent and cheap. The original text is always kept alongside the translation — nothing is overwritten, so you can always see exactly what the customer wrote. If a translation genuinely can't be produced, the system reports that and sends the original rather than shipping a mangled reply and calling it translated.
Ready to eliminate per-agent fees?
Start on Standard $49/mo for a complete modern helpdesk, or go Pro $99/mo for an autonomous AI resolution agent and agent copilot. Unlimited agents on both, flat per-workspace — no seat tax, no per-resolution AI billing. 14-day free trial, 20% off annual.