Aiolum Browse stories
All stories
On-screen AI training for staff of insurance agencies inside EZLynx and AMS360 · Canada

AgenQ: Product, team and how it was built

Solo founderFundedB2BInsurance

AgenQ (On-screen AI training for staff of insurance agencies inside EZLynx and AMS360, Canada): Live with paid plans for EZLynx and AMS360 users, November 2026. Evidence grade E: no revenue figure was checked. Team: 1 founder named (Varun Mishra, founder and CEO); headcount not published. Launched July 2025; still running.Added 7 November 2026

Outcome
Live with paid plans for EZLynx and AMS360 users, November 2026
Evidence
E no revenue figure was checked
Launch
July 2025
Status
Still running
Team
1 founder named (Varun Mishra, founder and CEO); headcount not published
Outside funding
The company lists NEXT Canada (NEXT AI), MaRS, MARL and NEC X as backers; no amount published, not confirmed
Main channel
Not stated
No revenue figure

What is not verified

No revenue figure was found. It was looked for on the company's pages, in Internet Archive copies of them, in the UpNext recap and in the local TrustMRR list of profiles, which has no AgenQ entry; no Canadian company register was searched. The company publishes no customer count and names no customer. The only result, about 90 minutes to about 15 for customer onboarding, comes from one confidential insurance software provider and is "customer reported"; UpNext repeated it as a company claim. The founding date and the headcount are not published. A reader could look for a named agency customer, a payment-processor profile or a register entry. This story is grade E: the business is real and its product, price and customers are described from public pages, but no revenue figure was checked. Aiolum shows it as an idea worth studying, not as proof that it makes money.

AgenQ is a Toronto company that sells in-app AI training: an assistant that runs on top of a business program and shows a user each click of a task on the screen, while the user does the real work. It works with two agency management systems, the programs in which insurance agencies keep their clients, policies and renewals: EZLynx and Vertafore AMS360. Staff add a Chrome extension and ask by typing or speaking; in "Execution Mode" the assistant can also complete approved tasks, with a person approving the important steps.

Varun Mishra, the founder and CEO, worked 15 years in enterprise software, including at MediaKind (formerly Ericsson Media), according to the company's about page. The first archived page of the product, of 15 July 2025, sold an "AI Product Assistant" to software companies; in February 2026 the assistant was called NINA, and by November 2026 the site sold it to insurance agencies. The pricing page, read on 7 November 2026, lists plans of US$199 and US$999 a month and a 7-day free trial.

No revenue figure was found, so this is a grade E story. The only result the company publishes comes from one unnamed insurance software provider, which it says cut customer onboarding from about 90 minutes to about 15; the company itself writes that this is not a result from an EZLynx or AMS360 agency. Varun Mishra pitched AgenQ among 45 AI startups at the SAAS NORTH AI PitchFest in Ottawa in November 2026, UpNext reported.

AgenQ: Live with paid plans for EZLynx and AMS360 users, November 2026. On-screen AI training for staff of insurance agencies inside EZLynx and AMS360, Canada, solo founder, evidence grade E.
The story in one card. Share it with the address of this page.
01

How much revenue does AgenQ make?

Each row shows the evidence grade and the source numbers from the list at the end of this page. The grades are explained in How Aiolum verifies.

A Toronto company that sells on-screen AI training for two insurance agency systems, with published prices and no published revenue figure
FactValueGradeHow we knowSources
Status, 7 November 2026live, with three published plans and a 7-day free trialCPricing page read on 7 November 2026: Starter US$199 a month (US$159 billed yearly), Growth US$999 a month (US$799 billed yearly), Enterprise custom.[2]
Supported softwareEZLynx and Vertafore AMS360CHome page and FAQ: "EZLynx and Vertafore AMS360 today, with more insurance software coming." The company states it is not affiliated with Applied Systems, EZLynx or Vertafore.[1, 6]
How it is delivereda Google Chrome extension; nothing installed inside EZLynx or AMS360CFAQ. Setup takes "a few hours of your agency's time"; the AgenQ team sets up the tasks.[6]
FounderVarun Mishra, founder and CEOCAbout page and press kit: 15 years in enterprise software, including MediaKind, formerly Ericsson Media, across engineering, pre-sales, sales and product; "Repeat founder".[3, 4]
Teamheadcount not publishedCThe about page and the press kit name only the founder. Some blog posts are signed "AgenQ Team". Offices: Toronto and Montreal.[3, 4, 7]
Backers listed by the companyNEXT Canada (NEXT AI), MaRS, MARL, NEC XCAbout page and press kit, with logos on the home page. No amount or date is given. A search of the NEXT Canada site on 7 November 2026 returned no page about AgenQ. Not confirmed.[1, 3, 4, 12]
First archived product page15 July 2025: "AgenQ - AI Product Assistant"CInternet Archive copy. Earlier captures of the domain are a Bluehost parking page (2017) and redirects (2019, 2021, 11 July 2025). No founding date is published.[9]
Product name, February 2026NINA, an AI product assistant for software companiesCInternet Archive copy of 19 February 2026: "Turn Your Software Into Its Own Expert with NINA".[10]
Only published customer resultonboarding about 90 minutes to about 15, about 30% fewer support callsCCase study published 27 May 2026, updated 25 October 2026. The customer is a North American insurance software provider, "name confidential", with hundreds of brokers. The company writes: "These are provider results, not measured results from an EZLynx or AMS360 agency deployment."[5]
Agency customersnot named; "a small number of agencies"CCase study: "We're working with a small number of agencies to set up AgenQ on the tasks their teams ask about most."[5]
SAAS NORTH AI PitchFest, Ottawa, November 2026one of 45 AI startups that pitched; not among the finalistsCUpNext recap of 6 November 2026, by Edwin J. Frondozo, lists AgenQ under "The Field" and repeats the company's onboarding figure: "The company says...". Varun Mishra pitched.[11]
Time from first archived product page to November 202616 monthscalcCalculated: July 2025 to November 2026.[9, 2]

How to read this table: grade C is a statement by the company on its own pages, an Internet Archive copy of those pages, or a report that repeats the company's words. A "calc" value was computed on this page. No row is a revenue figure.

AgenQ prices on its pricing page, read 7 November 2026 (grade C)
PlanPer month, billed yearly (US$)Per month, billed monthly (US$)Limit
Starter1591991 Product Expert, up to 3 guided tasks, 300 minutes of usage a month, self-service setup
Growth7999992 Product Experts, up to 15 guided tasks, 1,500 minutes of usage a month, assisted setup
Enterprisecustomcustomunlimited Product Experts, tasks and usage; private deployment; support with SLA

The page says the price depends on the number of "Product Experts" (assistants), guided tasks and usage minutes, not on the number of users. Usage above the monthly allowance costs US$0.07 a minute on Starter and Growth. Yearly billing saves 20%. Every plan has a 7-day free trial with no credit card; the buttons on the plans say "Talk to Sales".

02

When did AgenQ launch, and how did it grow?

From an AI assistant for any software company in July 2025 to a training tool for two insurance systems in November 2026. Each entry carries its grade.

  1. 15 July 2025CFirst archived product page: an "action-based AI product assistant" for software companies that demos, trains and executes inside their product, with "No credit card required". [9]
  2. 19 February 2026CThe archived home page calls the assistant NINA and sells it to software companies for onboarding and feature adoption. [10]
  3. 13 March 2026CA blog post describes Dashboard V2, the control panel for every deployed assistant, built on Next.js and React 19. [7]
  4. 27 May 2026CCase study: an unnamed insurance software provider cuts customer onboarding from about 90 minutes to about 15. [5]
  5. 25 and 26 October 2026CBlog posts on onboarding a new customer service representative in EZLynx and on agencies buying AI software; the case study, updated on 25 October, now ends with a section on EZLynx and AMS360. [5, 7]
  6. November 2026CVarun Mishra pitches AgenQ at the SAAS NORTH AI PitchFest in Ottawa; the pricing page lists three plans on 7 November. [11, 2]
03

Why did AgenQ work?

Four causes, in the order they mattered. Each one is a reusable lesson.

01

Help on the real screen

The assistant answers on the screen where the work happens and points to the next field or button, one step at a time. The company contrasts this with videos and guides, which "teach before or away from the task".

02

Two modes, with approval

Training Mode shows the steps and the person does them. Execution Mode does the task, and a person approves the important steps. The agency chooses who may use it, which tasks it may complete, and can read every step in the audit logs.

03

One industry, two systems

By moving from any software company to EZLynx and AMS360 only, the same tasks, such as updating a client's policy, repeat across many agencies. The company says setup takes a few hours of the agency's time, while the AgenQ team does the setup work.

04

Priced by tasks and minutes, not by seats

The pricing page lists plans by number of assistants, guided tasks and usage minutes, so an agency pays the same however many staff use it. The Starter plan lists self-service setup; Growth and Enterprise add assisted setup and a dedicated manager.

How it was built

These details come from the company's own pages and from Internet Archive copies of them. They are grade C. [6, 7, 2, 8]

  • Delivered as a Google Chrome extension that staff add to their browser; nothing is installed inside EZLynx or AMS360.
  • Customers manage their assistants in a web dashboard called Studio; the blog post of 13 March 2026 says Dashboard V2 is built on Next.js and React 19, with persona settings and voice integration per assistant.
  • Each guided task is "a process that AgenQ has learned"; the AgenQ team records and sets up the tasks with the agency.
  • A software development kit (SDK) for software companies that want AgenQ inside their own product is listed on every plan; on 7 November 2026 all of its documentation pages were marked "Soon".
  • No downloaded page names the language model, the speech recognition service or the voice provider.

Tools named in the sources

Each one opens its page in the Tools tab, with every story that uses it.

04

Who tried this before AgenQ, and why did they fail?

Three companies in our database of failed startups promised to help people at the moment they were stuck in a task or in software, and failed. The values below are from that database and are not independently verified.

Died 2010 to 2019Burned US$275M

Daqri

The promise

Step-by-step instructions and sensor data shown in an industrial worker's field of view through an augmented-reality helmet.

Why it died

The helmet was expensive and heavy, most customers lacked the IT systems and the will for deep changes to their workflows, and Microsoft competed with its own headset.

Product/Tech Failure
Died 2012 to 2015Burned US$120K

Hipmob

The promise

Live chat and in-app messages for customer support inside mobile apps and websites.

Why it died

Intercom and Zendesk added the same chat features, and Hipmob could not give customers a reason to switch.

Competition
Died 2013 to 2016Burned US$120K

Watchsend

The promise

Video recordings of how users moved through mobile apps, so that designers could see where users got stuck.

Why it died

Large analytics products such as Mixpanel and Google Analytics dominated, video data was costly to process, and developers did not want another SDK in their code.

Competition

The failed startups come from our database of 1,749 dead startups, retold here in our own words. Its entries are AI-assisted summaries of public sources, so these cards carry no evidence grade.

What changed

The three predecessors tried to help people at the moment they were stuck: Daqri with a helmet, Hipmob with a chat window, Watchsend with recordings for the app maker. Daqri needed costly hardware and deep changes to customers' systems, and Hipmob and Watchsend sold a feature that bigger platforms then added themselves.

AgenQ differs in two ways. Software that understands a typed or spoken question in the user's own words, delivered as a Chrome extension, needs no hardware and no change to the insurance system. The risk that remains is Hipmob's: the makers of the two systems can add the same guidance, and AgenQ's own FAQ already compares it with Vertafore Velocity Coach Agent.

05

Can one person copy AgenQ now?

Yes, for someone who knows one industry's software.

The method needs little capital: choose a program that a whole industry uses and that new staff find hard, set up guidance for the tasks people ask about most, and sell it to the firms that use the program. The part a copier cannot skip is selling to those firms; AgenQ's own case study says it was working with a small number of agencies in late 2026.

Where it is still open: Other industry programs with heavy training needs: practice management for dental and medical offices, property management, freight brokerage and public-sector case management.

  1. 1Choose one industry program with many screens and frequent new hires, and list the tasks that staff ask colleagues about most.
  2. 2Deliver the guide as a browser extension, so that nothing has to be installed in the vendor's system.
  3. 3Start with a mode that only shows the steps; add a mode that does the task only with a person's approval and an audit log.
  4. 4Price by tasks and usage minutes rather than by seats, so that a small firm can start.

What could go wrong for AgenQ?

The risks are the mirror image of the causes above.

What is not verified

No revenue figure was found. It was looked for on the company's pages, in Internet Archive copies of them, in the UpNext recap and in the local TrustMRR list of profiles, which has no AgenQ entry; no Canadian company register was searched. The company publishes no customer count and names no customer. The only result, about 90 minutes to about 15 for customer onboarding, comes from one confidential insurance software provider and is "customer reported"; UpNext repeated it as a company claim. The founding date and the headcount are not published. A reader could look for a named agency customer, a payment-processor profile or a register entry.

Trust signals on the site

Under "Backed by" the home page shows five logos (Marl Accelerator, MaRS, NEC X, NEXT AI, NEXT Canada), and the about page and press kit list NEXT Canada (NEXT AI), MaRS, MARL and NEC X, with no link, date or amount. A search of the NEXT Canada site on 7 November 2026 returned no result for AgenQ; the other backers were not checked. The case study names its customer only as "name confidential".

The product and its market changed

The July 2025 page sold an assistant to software companies with a "Human like Persona" described as "Realistic face, voice, and tone for natural interactions with personal name"; the November 2026 pages sell training to insurance agencies and show no face persona. The customer stories page is still headed "Built for SaaS". In Execution Mode the assistant acts inside EZLynx and AMS360, whose makers it says it is not affiliated with; the terms of those systems were not checked for this page.

Competition and model providers

No other story on this site sells on-screen training for one industry's software; the nearest insurance stories, Parsewise (insurance document packages) and LunaBill (insurance claim follow-up calls), solve different tasks, while AgenQ's own FAQ compares it with Vertafore Velocity Coach Agent and ChatGPT. No downloaded page names the AI model, speech or voice provider, so the dependence on model providers cannot be stated.

Sources

The grade next to each source is the best grade that the source can support. "Read in full" means that the text of the page was downloaded and the numbers were checked in it.

  1. C AgenQ home page: AI Training and Onboarding for Insurance Software
    Read in full on 7 November 2026. Source for the product, the two modes, the controls, the "Backed by" logos and the questions section.
  2. C AgenQ pricing page
    Read in full on 7 November 2026. Source for the plans, prices, usage limits, overage price and free trial.
  3. C About AgenQ
    Read in full on 7 November 2026. Source for the founder, his background, the listed backers and the two offices.
  4. C AgenQ press kit
    Read in full on 7 November 2026. Confirms the founder, the city, the supported software and the listed backers.
  5. C AgenQ, 27 May 2026, updated 25 October 2026: How AgenQ Cut Insurance Software Onboarding Time by 6X
    Read in full on 7 November 2026. The company's case study; source for the only customer result and its limits, and for the small number of agencies.
  6. C AgenQ frequently asked questions
    Read in full on 7 November 2026. Source for the Chrome extension, the setup time, the supported systems, the affiliation statement and the comparison with Vertafore Velocity Coach Agent.
  7. C AgenQ blog index
    Read in full on 7 November 2026. Source for the post dates, the authors and the Dashboard V2 post of 13 March 2026.
  8. C AgenQ documentation
    Read in full on 7 November 2026. Every article, including the SDK installation, is marked "Soon".
  9. C Internet Archive copy of agenq.com, 15 July 2025: AgenQ - AI Product Assistant
    Read in full. The first capture with product content; the capture list also shows a 2017 parking page and redirects before July 2025.
  10. C Internet Archive copy of agenq.com, 19 February 2026: Home - AgenQ
    Read the opening sections. Source for the NINA name and the focus on software companies.
  11. C UpNext, 6 November 2026, by Edwin J. Frondozo: SAAS NORTH AI 2026: DOUBL Wins PitchFest as 45 AI Startups Pitch Ottawa
    Read the introduction, the finalists and the entry for AgenQ under "The Field". Independent confirmation that Varun Mishra pitched AgenQ; the onboarding figure is repeated as a company claim.
  12. C NEXT Canada site search for AgenQ
    Read on 7 November 2026. The search returned "Nothing Found"; used only to say that the listed backing was not confirmed.
  13. C Database of failed startups, local export
    Local export. AI-written summaries, not independently verified. Source for the failed predecessors.

Run this company? Send corrections or a verified figure to aiolum@livyden.com

Back to all stories