Curious Thing Review

Curious Thing AI sells named voice AI agents — Lucy, Alex, Sam & Lou — that make and take business phone calls: answering missed calls, handling FAQs and appointment bookings, qualifying leads, and running outbound follow-up and payment-reminder campaigns. The company was recently acquired by contact-center technology firm MaxContact, and its public pricing page has since disappeared. This review covers what the product does today, what independent evidence exists, and what changed around the acquisition.

Verdict at a glance

What kind of tool is Curious Thing?

Curious Thing AI builds conversational voice agents for phone-based customer interactions. Rather than one generic chatbot, it ships four named products: Lucy answers missed inbound calls and texts back a summary; Alex handles FAQs, qualifies leads, and manages appointment booking and rescheduling with call routing; Sam & Lou make outbound calls for follow-ups, payment reminders, and chasing missing documents; and a Custom tier wraps any of the above to ISO 27001, SOC 2, GDPR, and HIPAA standards for regulated industries. Customers named on the site — Brighte, InDebted, and Quitline, all Australian organizations — each gave their deployed agent a first name (Kate, Brooke, Matilda), the product's signature move: the AI is presented as a team member, not a bot.

As of mid-2026, the homepage carries a banner stating the company was acquired by customer-engagement technology firm MaxContact, and the footer now credits the product to MaxContact too — a meaningful change worth knowing before you buy, since you'd be dealing with a different company than the one that built the original product's reputation.

How it works

Deployment is agent-based rather than seat-based: you pick Lucy, Alex, Sam & Lou, or a custom build, and Curious Thing configures the conversation flows, voice, and — per its marketing — a distinct personality for your brand. The vendor advertises multilingual and multichannel support (text and email alongside the call), CRM and call-center integrations, an API, and real-time transcripts and call summaries. None of this is independently benchmarked; it's vendor-described functionality, not something we tested.

The vendor also cites its own usage figures: 3 million minutes of AI-human conversation processed in the past 12 months, 82% of end users reporting a positive experience, and one campaign reaching 1200% ROI on membership renewals. These are self-reported marketing claims with no stated methodology, so treat them as directional rather than verified.

Pricing

Curious Thing no longer publishes a pricing page — navigating to it redirects to the homepage — and Capterra lists its pricing model as usage-based, starting price "contact vendor." The one concrete number available is secondhand: a June 2025 Capterra reviewer said their Professional plan had been raised to $24/month, and that a previously free version was discontinued around when the reviewer says a company called Trivoni Software took over. That name doesn't match the MaxContact acquisition banner on the current site, and we could not reconcile the two — it may reflect an earlier ownership change, a rebrand, or reviewer error. Treat any pre-2026 pricing figure as unconfirmed and get a current quote before assuming it still applies.

What we could and couldn't verify

Independent review volume here is very thin. Curious Thing's Capterra listing shows an overall 3.0/5 from exactly two reviews: a 2021 five-star review praising the product for a recruitment and candidate-experience use case, and a 2025 one-star review titled "This used to be a great company and experience," describing unanswered support emails and a discontinued free tier after an ownership change. G2 and Trustpilot's review pages both loaded only bot-verification walls, so neither could be checked, and we found no active Reddit discussion of the product. That leaves vendor claims and testimonials on the company's own site as the bulk of available evidence — a gap we're flagging rather than papering over.

Pros and cons

Pros

• Purpose-built products (Lucy, Alex, Sam & Lou) for distinct inbound and outbound use cases, rather than one generic bot
• Compliance-oriented Custom tier (ISO 27001, SOC 2, GDPR, HIPAA) for regulated buyers
• Named, brand-customizable agents with real-time call summaries and multilingual support

Cons

• No public pricing; you must go through a sales conversation to learn cost
• Extremely small independent review base (2 Capterra reviews) — not enough to judge reliability
• A documented 2025 complaint about support responsiveness and a discontinued free tier following an ownership change
• A recent acquisition (MaxContact) makes near-term roadmap and support continuity uncertain

Curious Thing alternatives

Voice AI phone agents are a fast-moving category with contact-center incumbents and newer AI-native entrants:

• Five9 — established cloud contact-center platform that has layered AI agents onto mature, enterprise-grade call infrastructure.
• PolyAI — enterprise-focused voice AI with a track record in large-scale customer service deployments.
• Synthflow AI — self-serve voice agent builder aimed at smaller teams that want to configure agents without an enterprise sales cycle.
• Bland AI — developer-oriented voice AI platform with an API-first approach for custom-built calling agents.
• Retell AI — API and SDK-based voice agent infrastructure popular with teams building their own calling products on top.

Who should use Curious Thing — and who shouldn't

Good fit: businesses in regulated or compliance-sensitive industries (financial services, healthcare, fintech) that want a vendor already positioning itself around ISO 27001, SOC 2, GDPR, and HIPAA; teams that specifically want separate, named agents for inbound versus outbound work rather than one general-purpose bot.

Poor fit: buyers who want transparent, self-serve pricing before talking to sales; anyone uneasy adopting a product mid-acquisition, where support and roadmap continuity aren't yet proven under the new owner; teams that need a real body of independent reviews to de-risk the decision — that evidence doesn't exist yet here.