> ## Content Index > Fetch the complete content index at: https://www.theaidirectory.com.au/llms.txt > Use this file to discover other available public pages before exploring further. # Lorikeet Review (2026): The Sydney-Built AI Support Agent for High-Stakes Tickets - URL: https://www.theaidirectory.com.au/review-lorikeet-australia-2026/ - Published: 2026-06-12T23:00:00.000Z - Updated: 2026-10-09T06:32:40.000Z - Description: A Sydney-founded AI support agent built to resolve complex fintech and healthcare tickets, not deflect them. Strong product, US-dollar pricing, and one residency gap. - Author: Adam - Tags: Customer Service & AI Chatbots, Lorikeet, #review **20**/25 ★★★★☆ Very Good Lorikeet Verified 21 Sep 2026 Reviewed 13 Jun 2026 Value for Money**4**/5 Ease of Use**3**/5 Feature Depth**5**/5 Australian Fit**4**/5 Reliability & Support**4**/5 From **A$2,100**/mo · ex-GST · billed in USD **The gate:** Five stars need 21+ and Australian Fit of at least 4/5. [Skip to the verdict ](#verdict) > **The verdict:** Lorikeet is one of the few genuinely Australian-founded AI products operating at the serious end of customer support. It is built to *resolve* complicated, high-trust tickets — verify a customer, check a backend system, take an action — not just deflect them to a help article. The product is excellent. The catch for Australian buyers is that pricing is in US dollars and there is no published Australian data-residency option. For a fintech or healthtech support team that can wear those two things, it is one of the strongest tools in this category. For a small Melbourne business answering 200 tickets a month, it is overkill. **Score: 20/25 — Very Good.** ## What it is Lorikeet is an AI customer-support platform. Strip away the marketing word "concierge" and here is what it actually does: it sits in front of your support queue and tries to resolve tickets end to end, the way a trained human agent would. That distinction matters. Most "AI chatbots" in this category are retrieval bots. They read your help centre, summarise an article, and hope that answers the question. Lorikeet is built to do the next step — authenticate the customer, look up their account in a backend system, perform an action (issue a refund, reset a setting, check a claim), and then explain the result over chat, email, SMS or voice. It was founded in Sydney in 2023 by Steve Hind (ex-Stripe) and Jamie Hall (ex-Google Brain, a co-author on Google's LaMDA language-model paper). The company is genuinely Australian in origin, though it now runs a team across Sydney and the United States and sells globally. Its public customer stories lean heavily on fintech and healthtech — industries where a wrong answer is expensive and the rules are strict. In plain terms: this is not a tool for putting a friendly bubble on your website. It is a tool for replacing the slow, repetitive parts of a real support operation. ## What it costs in Australia (AUD + GST) Lorikeet publishes its pricing, which is rare in this category and worth crediting. It is **billed in US dollars** and runs on a **pay-per-resolution** model — you are charged when a ticket is actually resolved, not for seats or for every message. The two published tiers, with an approximate AUD conversion at \~1.40 (see GST note below): | Plan | Price (USD) | Approx. AUD/month | Best for | | -------------- | -------------------------- | ----------------- | --------------------------------- | | **Start** | US$1,500/mo (US$18,000/yr) | \~A$2,100/mo | Under 5,000 tickets/month | | **Scale** | US$4,000/mo (US$48,000/yr) | \~A$5,600/mo | 5,000–20,000 tickets/month | | **Enterprise** | Custom quote | — | 20,000+ tickets or complex setups | On top of the platform fee, resolutions are charged per ticket and vary by channel: roughly **US$0.95 down to US$0.80** for chat, email and SMS, and **US$1.50 down to US$1.20** for voice, with the lower numbers on the Scale tier. Routing, tagging and QA-only actions are cheaper, around US$0.25–$0.30 each. There are no per-seat charges and no implementation or platform setup fee. **GST treatment:** Because the entity invoicing you is likely a US-based one charging in USD, this is most often a B2B import of services. Under Australian rules an Australian GST-registered business generally accounts for GST itself (reverse charge) rather than paying it on the invoice, and can usually claim it back — so the *effective* cost to a registered business is close to the AUD figures above, not those figures plus 10%. Lorikeet does not state on its pricing page whether it is registered for Australian GST or which entity issues the invoice — confirm with their sales team and your accountant before budgeting. The honest read: this is mid-market-to-enterprise pricing. The US$1,500/month floor (about A$2,100) means the maths only works once you have real ticket volume. A sole trader or a five-person business will not get value here. ## How it works / features Lorikeet's core idea is an "agentic" architecture rather than a single prompt-and-reply bot. A few specifics worth knowing: - **It takes actions, not just answers.** Through integrations it can read and write to your systems — look up an order, update a record, trigger a workflow — so it can close a ticket rather than route it. - **It runs on a graph-based control plane.** Instead of letting the model freewheel, Lorikeet maps your support procedures as a structured graph of steps and guardrails. This is the feature that makes regulated buyers comfortable: behaviour on a sensitive workflow is predictable, not improvised. - **It is multichannel.** Chat, email, SMS and voice are all supported, billed at the per-channel rates above. - **It plugs into the helpdesks you already run.** Published integrations include Zendesk, Intercom, HubSpot, Front and Salesforce, so it layers onto an existing stack rather than replacing it. - **You only pay for resolutions.** Lorikeet's stated policy is that if you are unhappy with how it handled a ticket, you do not pay for that ticket. That is an unusually buyer-friendly stance and signals confidence in the resolution quality. The throughline is trust. Every design choice — the graph control plane, the resolution-only billing, the focus on regulated industries — points at the same buyer: a support leader who cannot afford the bot to go off-script. ## What it does well **It resolves, it doesn't deflect.** This is the single biggest reason to look at Lorikeet over a cheaper retrieval chatbot. Connecting to backend systems and taking real actions is what separates "deflection theatre" from genuine first-contact resolution. **The guardrails are the product.** For fintech, healthtech, insurance and similar, the graph-based control plane is a serious differentiator. It is the answer to the obvious objection — "I am not letting an AI loose on my customers' accounts." **Pricing is published and aligned.** Charging only for successful resolutions ties Lorikeet's revenue to your outcomes. In a category full of opaque "contact sales" pricing, putting real numbers on a public page earns trust. **The pedigree is real.** Stripe and Google Brain founders, backing from Blackbird, Square Peg, Skip Capital and QED, and roughly US$75 million raised across 2024–25\. This is not a weekend project that will vanish in a year. ## Where it falls short **It is built in USD, and that flows downhill to you.** Pricing, contracts and the per-resolution rates are all dollar-denominated. Your monthly bill will move with the exchange rate, which makes budgeting harder than a tool that simply quotes AUD. This is the clearest mark against its Australian fit. **No published Australian data residency.** This is the one I would press hardest on. Lorikeet's public material lists SOC 2 Type 2 and ISO 27001, with a HIPAA BAA available on the Scale tier — strong credentials. But I could not find any statement that customer data can be hosted in an Australian region. For an Australian fintech, health provider or government-adjacent buyer with data-sovereignty obligations, that is a procurement question to settle in writing before signing, not after. **It is too big for small business.** The US$1,500/month floor and the focus on complex, high-volume support make this the wrong tool for most of our readers. If you are a café, a tradie or a boutique agency, the value is not there — you want a lighter, cheaper option from our category roundup. **You have to do the work to make it sing.** Mapping your procedures into the control plane and wiring up backend integrations is a project, not an afternoon. The payoff is real, but so is the setup cost in time and engineering. ## Australian fit This is the interesting part, because Lorikeet pulls in two directions. **On origin: strongly Australian.** Founded in Sydney, still operating a Sydney team, backed substantially by Australian venture funds. If "support a local company" matters to you, Lorikeet is one of the few names in this category that genuinely qualifies. It is a rare Australian AI export success story. **On the practical buyer experience: only partly Australian.** Billing is in US dollars, not AUD. There is no published AU-region data-residency option. GST handling on the invoice is unstated. Support is global rather than explicitly Australian-hours. So while the *company* is Australian, the *product as sold* behaves like an international SaaS tool — which is a different thing from the local-billing, local-hosting fit we prize for everyday Australian businesses. Per our methodology, Australian Fit measures the buyer's experience — AUD billing, data residency, GST, AU support and AU-aware outputs — not the founder's postcode. On that test Lorikeet scores solidly but not at the top, held back by USD pricing and the residency gap. Under our Australian Fit gate, that keeps it from a five-star or category-winner rating despite a strong raw score. ## Score | Criterion | Score | Notes | | --------------------- | --------- | ---------------------------------------------------------------------------------------------------- | | Value for Money | 4/5 | Resolution-only billing and published pricing are excellent; the US$1,500 floor limits who benefits. | | Ease of Use | 3/5 | Powerful, but mapping procedures and wiring integrations is a real project. | | Feature Depth | 5/5 | Takes actions, multichannel, graph-based guardrails — genuine end-to-end resolution. | | Australian Fit | 4/5 | Sydney-founded and AU-backed, but USD billing and no published AU data residency hold it back. | | Reliability & Support | 4/5 | SOC 2 Type 2, ISO 27001, HIPAA BAA on Scale, serious backing; AU-hours support unstated. | | **Total** | **20/25** | | ⭐⭐⭐⭐ **Very Good** ## Bottom line **If you run a fintech or healthtech support team in Australia:** this should be on your shortlist, possibly at the top of it. The action-taking, the graph guardrails and the regulated-industry posture are exactly what you need. Settle the data-residency and GST questions in writing first, then trial it against your real ticket mix. **If you run a mid-sized business with genuine ticket volume (say 3,000+ a month) and an engineering resource to set it up:** worth a serious look. The resolution-only pricing means you are paying for outcomes, which de-risks the trial. **If you are a small Australian business:** skip it for now. The floor price and setup effort are wrong for you. Start with a lighter, AUD-billed option — see the roundup below. Update history - **13 Jun 2026** — First reviewed. - **21 Sep 2026** — **Verified by browser from an Australian IP: Lorikeet still prices in US dollars — notable, because Lorikeet is an Australian company.** lorikeetcx.ai/pricing names USD and shows US$2,100 and US$5,100 platform fees with per-resolution rates in US dollars. A Sydney-built product billing Australians in US dollars is worth saying out loud rather than leaving implied. No change to the review's currency position; `aud_native: false` is now set explicitly in the frontmatter so the directory stops inferring it. Source: https://www.lorikeetcx.ai/pricing · Wayback: capture pending (archive.org returning 500/429 since 19 Sep 2026). Checked by the `aud-recheck-l-to-z-2026-09-21` pass. ## Frequently asked questions ### Is Lorikeet an Australian company? Yes — genuinely. Lorikeet was founded in Sydney in 2023 by Steve Hind (ex-Stripe) and Jamie Hall (ex-Google Brain). It still runs a Sydney team and is backed by Australian funds including Blackbird, Square Peg and Skip Capital. The catch: the product as sold bills in US dollars and behaves like an international SaaS tool. ### What does Lorikeet cost for Australian businesses? Lorikeet bills in US dollars on a pay-per-resolution model. The Start tier is US$1,500/month (about A$2,100) for under 5,000 tickets a month. Scale is US$4,000/month (about A$5,600) for 5,000–20,000 tickets. Resolutions cost roughly US$0.80–$0.95 each for chat, email and SMS, more for voice. There are no per-seat charges and no setup fee. ### What score did Lorikeet get in this review? Lorikeet scored 20/25 — Very Good, four stars. Feature Depth earned a perfect 5/5 for its action-taking, multichannel support and graph-based guardrails. Australian Fit landed at 4/5: Sydney-founded and AU-backed, but USD billing and no published Australian data residency hold it back from the top rating. ### Does Lorikeet store customer data in Australia? There is no published Australian data-residency option — the biggest caution in this review. Lorikeet holds SOC 2 Type 2 and ISO 27001, with a HIPAA BAA available on the Scale tier. If your organisation has data-sovereignty obligations, settle the residency question in writing before signing, not after. ### Who should use Lorikeet — and who shouldn't? Australian fintech and healthtech support teams should shortlist it, possibly at the top. Mid-sized businesses with 3,000-plus tickets a month and an engineering resource should take a serious look. Small businesses should skip it: the US$1,500/month floor and real setup effort make a lighter, AUD-billed option the better buy. ## Read next - [Our review methodology and scoring framework](https://www.theaidirectory.com.au/methodology/) — how we score, and what the Australian Fit gate means. - [Best AI Customer Service Tools for Australian Businesses (2026)](https://www.theaidirectory.com.au/best-ai-customer-service-chatbots-australia/) — the full category roundup, including lighter and cheaper options for smaller teams. ## Sources Primary sources, checked 13 June 2026 (Wayback snapshots queued): - Lorikeet — About / company page: https://www.lorikeetcx.ai/about - Lorikeet — Pricing (tiers, per-resolution rates, SOC 2 / ISO 27001 / HIPAA BAA): https://www.lorikeetcx.ai/pricing - Lorikeet — Series A announcement (US$35M, QED Investors): https://www.prnewswire.com/news-releases/lorikeet-raises-35-million-usd-series-a-to-deliver-ai-powered-universal-customer-concierges-able-to-solve-real-challenges-not-just-chatbots-302522821.html Secondary / context: - Startup Daily — "Lorikeet has now raised $75 million in just 10 months": https://www.startupdaily.net/topic/funding/ai-customer-support-startup-lorikeet-has-now-raised-75-million-in-just-10-months/ - AAP — Lorikeet secures additional A$9M (Blackbird, Square Peg, Skip Capital): https://aap.com.au/aapreleases/cision20250204ae11951/ - Built In Sydney — Lorikeet Sydney office profile (confirms Sydney base): https://builtinsydney.au/company/lorikeet **Structured data (Ghost code injection):**