In 2015, two software people in Bozeman, Montana had an idea that sounded almost too small to build a company on: what if talking to a business felt like texting a friend? No hold music, no "your call is important to us," no repeating your account number four times. Just a thread you could pick up whenever. A decade later that small idea has grown into Quiq - an enterprise AI platform that quietly runs customer-service agents in production for more than 150 global brands, including IHG Hotels & Resorts, Volvo, HBO, Lululemon and Roku.
The company's founders, Mike Myer and William O'Neill, were not exactly newcomers when they started. Myer had helped scale Dataminr and sat on the board of CradlePoint, a networking company Ericsson later bought for $1.1 billion. He could have built almost anything, almost anywhere. He built a messaging company in a ski-and-trout town most enterprise buyers would struggle to find on a map. Ten years on, that choice looks less like a quirk and more like a strategy.
It also explains the pace. Companies founded on a coast tend to raise fast and burn faster, chasing the next headline. Quiq raised roughly $47.5 million total across a decade - restrained by the standards of its category - and used the time to do something less glamorous than a viral demo: learn what actually breaks when a real brand hands real conversations to software. That patience is not a personality trait. It is a product advantage, and it shows up in the parts of the platform customers never see.
01 / What Quiq Actually DoesAI agents, with the safety switches left on
Strip away the category language and Quiq builds AI agents that handle customer conversations from first hello to resolution. Its AI Agents answer inquiries directly, pulling from a brand's own knowledge base so the answers are the company's answers and not a chatbot's best guess. Its AI Assistants sit beside human agents, drafting replies and surfacing information so one person can juggle several conversations at once. And as of May 2026, its Voice AI extends all of that into spoken conversation, so a customer can start in chat and end on a call without repeating a word.
The current homepage line is blunt about the pitch: "Powerful AI agents with enterprise control you can trust." That second half is the whole company. Plenty of vendors can demo an impressive agent. The hard part - the part enterprises actually lose sleep over - is control: knowing what the agent will and won't say, proving it to a compliance team, and being able to hand a conversation to a human the moment it should be handed off.
02 / Who Uses ItBig brands, unglamorous wins
Quiq's customer list reads like an airport departures board: IHG Hotels & Resorts, Volvo, HBO, Lululemon, Panasonic, Roku, Brinks Home, Jos. A. Bank, Terminix, Progressive Leasing, Bob's Discount Furniture. These are companies with millions of customer conversations and very little appetite for surprises. The results they report are the kind that show up in an operations review rather than a keynote: Roku reduced cost by roughly 5x, and Brinks Home cut cost per contact by 67% while lifting customer satisfaction by double digits.
One detail says more about the platform than any logo wall. According to the company's May 2026 announcement, a single global retailer runs one Quiq AI agent to support four different brands across seven countries and four communication channels. That is not a pilot bolted onto a landing page. That is one configuration doing the work of what used to be dozens of disconnected systems.
03 / The ProblemThe pilot graveyard
If you have followed enterprise AI at all, you know the pattern. A company runs a flashy AI pilot, it demos well, the press release goes out - and then it never ships to real customers at real volume. The gap between "it worked in the demo" and "it runs our support line on Black Friday" is where most customer-service AI quietly dies.
Quiq's entire 2026 rebrand is aimed at that gap. The company reframed its identity around AI's move from experimental tooling into operational infrastructure - the boring, load-bearing kind of software an enterprise actually depends on. It is a telling thing to build a brand around. Most companies want to sell you the magic. Quiq is selling the part that survives contact with a compliance review and a traffic spike.
04 / How It's DifferentModel-agnostic, on purpose
Most AI vendors marry a single model and hope it stays the best one. Quiq made the opposite bet: it is deliberately model-agnostic. Its Voice AI runs across ElevenLabs, Deepgram, Google and Amazon voice engines, and its agents run on multiple large language models rather than one. When a better model arrives - and in this field one always does - Quiq can swap it in without asking its customers to re-platform.
The second difference is control, and it is more than a slogan. Quiq's platform is SOC 2 Type II, HIPAA, GDPR and CCPA compliant, and it advertises 99.999% uptime - the sort of specification you only bother printing when the buyer is a bank, a hospital, or a hotel chain. The company positions the enterprise, not the algorithm, as the party holding the steering wheel. For a Terminix or an ICU Medical, that is the difference between a feature and a liability.
There is a third, quieter difference: continuity across channels. A customer who begins a question over SMS, gets pulled into a phone call, then follows up by email is, to most systems, three different strangers. Quiq's stated aim is to keep that one person as one conversation - the context traveling with them from chat to voice to inbox. It sounds like a small courtesy. For anyone who has ever re-explained the same problem to the same company three times, it is the entire point.
05 / The BusinessYou pay for conversations, not chairs
Quiq's pricing is quietly one of its most interesting choices. Instead of charging per seat - the default for a generation of customer-service software - it charges based on usage: you pay for the conversations and resolutions you actually use. In a world where AI agents are supposed to reduce the number of human seats you need, per-seat pricing would be self-defeating. Per-conversation pricing lines Quiq's incentives up with its customers': the more the AI resolves, the more the model earns, and nobody is paying for empty chairs.
The capital story matches the pricing story: measured. Quiq raised a $12.5M Series B in 2019 led by Foundry Group and a $25M Series C in 2022 led by Baird Capital, bringing total funding to roughly $47.5 million. By 2024, outside estimates put revenue near $20 million with a team of fewer than 100 people. In an industry where AI companies raise nine figures before they have a product, Quiq's numbers read like restraint - a company built to last a decade rather than a news cycle.
06 / Expertise & PositionA decade of talking to strangers
Quiq's real moat is time. It has been helping brands run digital messaging at scale since before "conversational AI" was a category anyone bragged about. That decade of watching real customers argue, complain, buy and cancel across SMS and chat is exactly the raw material an agentic platform needs to be trusted with the hard conversations - not the demo questions, the 2 a.m. ones.
In the market, Quiq sits among the agentic-CX players - Ada, Sierra, Intercom's Fin, Zendesk AI, Salesforce Agentforce, LivePerson, Kore.ai and Cognigy. What separates it is less a single feature than a posture. It won its 2024 recognition - Best Conversational Customer Service AI Platform at the World Future Awards, plus a CUSTOMER Magazine innovation award - not for the flashiest agent, but for making agents an enterprise could actually put in production and defend to a regulator. In a category still crowded with demos, that turns out to be the rarer skill.
There is a small irony worth sitting with. A company whose founder helped build a $1.1B networking business chose Montana, chose restraint, and chose to sell the least exciting promise in AI - reliability - to some of the most demanding buyers on earth. Ten years and 150 brands later, the unglamorous bet is the one still standing.