Haiket

Pioneering context-aware voice messaging for natural communication experiences.

Website: https://haiket.com/

Cover Block

PUBLIC

Field Value
Name Haiket
Tagline Pioneering context-aware voice messaging for natural communication experiences
Headquarters London, UK
Stage Pre-Seed
Business Model B2C
Industry Consumer Communications
Technology Type AI / Machine Learning
Geography Western Europe
Growth Profile Venture Scale
Founding Team Solo Founder (Alexander Narest)

Links

PUBLIC

Executive Summary

PUBLIC

Haiket is a London-based consumer messaging startup attempting to make voice, rather than text, the default surface for mobile conversation, using what it calls context-aware technology to interpret tone, environment and conversational state [PRNewswire, Nov 2025]. The company opened a public waitlist and invited European users to test the app on November 11, 2025, marking its first formal product disclosure [Webull, Nov 2025]. It was founded by Alexander Narest, a voice technology engineer whose prior work includes contributing to audio quality improvements in Google Duo via WaveNetEQ and AI-based packet loss concealment, and who previously built Holler, a voice messaging app with real-time translation across 65 languages [LinkedIn; AppAdvice]. The business model is consumer-facing and the company is at pre-seed stage; no funding rounds, investor names, or revenue figures are publicly disclosed at the time of writing. Differentiation, per the company's positioning, rests on context awareness layered on top of voice messaging rather than on a new transport or social graph [Haiket]. The product enters a category dominated by WhatsApp, iMessage and Telegram voice notes, which means execution risk centers on whether context-aware features generate enough perceived utility to pull users away from incumbent inboxes. Over the next 12 to 18 months, the signals worth tracking are waitlist conversion, any disclosed seed financing, partnership or distribution announcements, and whether the founder converts his Google and Fring network into early go-to-market support.

Data Accuracy: YELLOW -- Confirmed by PRNewswire and Crunchbase; founder background corroborated by LinkedIn and AppAdvice; financing and traction unverified.

Taxonomy Snapshot

Axis Value
Stage Pre-Seed
Business Model B2C
Industry / Vertical Consumer Communications / Messaging
Technology Type AI / Machine Learning (voice)
Geography Western Europe (London HQ)
Growth Profile Venture Scale
Founding Team Solo Founder

Company Overview

PUBLIC

Haiket surfaced publicly on November 11, 2025, when the company issued a press release announcing the European launch of a context-aware voice messaging application and opened a customer testing program [PRNewswire, Nov 2025]. The company is headquartered in London and lists itself on Crunchbase under the description "pioneering the next advancement in human-technology interaction, from screens to speech" [Crunchbase]. A founding date is not disclosed in any of the captured public sources, and the legal entity behind the brand has not been independently verified through Companies House records in this report.

The founding story, as told in the launch coverage, is a continuity narrative rather than a pivot. Alexander Narest, the founder and CEO, frames Haiket as the next chapter in a career that has moved between voice AI at Google and voice-over-IP work at Fring, the Israeli mobile VoIP company acquired by Genband in 2013. "From voice AI at Google to VoIP at Fring, my career has focused on building voice firsts in technology," Narest said in the launch announcement, adding that Haiket is intended to move "communication forward, with a new, voice-first" experience [Yahoo Finance, 2025].

Key public milestones to date are limited: the November 2025 European launch and the opening of a waitlist on the company's website, which currently presents only the line "Voice-first messaging, perfected. Join the waitlist" [Haiket]. The launch was accompanied by a supporting quote from John Donovan, who referenced his role in the iPhone launch on AT&T [PRNewswire, Nov 2025]; his formal relationship to Haiket (advisor, investor, board member) is not specified in the cited materials.

Data Accuracy: YELLOW -- Confirmed by PRNewswire and Crunchbase; founding date and legal entity not yet independently verified.

Product and Technology

MIXED

Haiket positions itself as a voice-first messaging application whose core differentiator is "context awareness," a phrase the company uses to describe technology that adjusts the voice messaging experience based on environmental and conversational signals to produce "a more natural experience" [PRNewswire, Nov 2025] [PUBLIC]. The published materials do not specify which signals the system uses (ambient noise, prior message history, speaker identity, language, device state) or how the inference runs (on-device versus cloud), and the company has not released a technical white paper at the time of writing [PUBLIC].

The product surface that is publicly visible is minimal: a marketing site at haiket.com inviting users to join a waitlist, and an Android listing reviewed by Softonic that describes Haiket as an AI chat application [Haiket; Softonic] [PUBLIC]. The European user testing program announced on November 11, 2025 is the first opportunity for outside parties to evaluate the app in operation [Webull, Nov 2025] [PUBLIC]. No App Store listing, SDK, API, or developer documentation has been surfaced in the captured sources.

On the technology side, the most credible signal is the founder's prior body of work. Narest's LinkedIn record references his contribution to "Improving Audio Quality in Duo with WaveNetEQ," the Google research project that used a generative neural network to fill in audio gaps caused by packet loss during VoIP calls [LinkedIn] [PUBLIC]. He has also shipped Holler, a consumer voice-and-text app with real-time translation across 65 languages [AppAdvice] [PUBLIC]. Both data points suggest the team has direct, hands-on experience with the audio-pipeline and machine learning components that a context-aware voice messenger would require, although it is not confirmed which of those technologies, if any, are present in the Haiket build [MIXED].

Data Accuracy: ORANGE -- Product claims rely on a single PRNewswire release and the company's own website; founder technical background is GREEN via LinkedIn and AppAdvice.

Market Research and Opportunity

PUBLIC

Voice as an interaction layer is having a second moment, driven by a step-change in generative speech models and a reset in how consumers expect to talk to software. Haiket is entering the consumer messaging category, where the dominant unit of communication has been text, supplemented by asynchronous voice notes that users still describe as awkward to record and tedious to consume.

No third-party TAM, SAM or SOM figure for context-aware voice messaging is cited in the captured research, and Haiket itself has not published a sizing claim. As an analogous reference, the global mobile messaging app category is dominated by WhatsApp, which Meta reported crossing two billion users in 2020 and which remains the default messenger across most of Western Europe (analogous market, public Meta disclosures, not cited in the structured facts and therefore presented only as directional context). The relevant addressable surface for Haiket in its launch geography is the subset of European smartphone users who already send voice notes inside an existing messenger and might adopt a dedicated voice-first app, plus users of voice assistants who want a more conversational asynchronous channel.

Demand drivers surfaced in the launch coverage are qualitative rather than quantitative. John Donovan, quoted in the launch release, framed the opportunity as the "next evolution of messaging, making voice the center of digital connection" and explicitly referenced WhatsApp as the prior platform shift [PRNewswire, Nov 2025]. The implicit thesis is that improvements in speech synthesis, transcription accuracy and on-device inference have made voice fast enough and accurate enough to be a primary interface rather than a secondary one. Adjacent and substitute markets that compete for the same user attention include in-messenger voice notes (WhatsApp, iMessage, Telegram), AI voice companions (such as those built on conversational LLMs), and dictation-plus-text workflows that convert spoken input into written output.

Regulatory context for a London-headquartered consumer communications app is non-trivial. The UK Online Safety Act 2023 imposes content moderation duties on user-to-user services, and the EU's Digital Services Act applies to any service offered to users in the bloc; both regimes require a voice messaging operator to think early about abuse handling, age assurance and lawful access requests. None of the captured Haiket sources address how the company plans to meet these obligations.

Sizing claim Value Source
Confirmed Haiket-cited TAM/SAM Not publicly available n/a
Launch geography Europe [PRNewswire, Nov 2025]
Launch date November 11, 2025 [Webull, Nov 2025]

The analyst takeaway: there is no defensible market-size number to anchor a Haiket model today. The investable thesis at this stage rests on category timing (voice as an interaction primitive) and founder fit, not on a sized opportunity, and any seed conversation will need a primary-research sizing exercise from the investor side.

Data Accuracy: ORANGE -- Market framing relies on a single press release and analogous public context; no third-party report on context-aware voice messaging has been cited.

Competitive Landscape

MIXED

Haiket is launching into a category where the default behavior already exists inside applications users open dozens of times per day, which means the competitive question is less about features than about whether voice-first justifies a separate inbox. The structured facts surface no named competitors, so the comparison below is drawn from publicly known incumbents and adjacent challengers rather than from cited Haiket research [PUBLIC].

The segment-by-segment map breaks into three groups. The first is the incumbent messenger group: WhatsApp, Apple iMessage, Telegram and Meta Messenger all support voice notes natively, distribute through the system address book, and have effectively zero acquisition cost for adding a new conversational mode. They are not voice-first, but they are voice-capable, and the friction cost of opening a separate app is high. The second group is voice-AI consumer apps, including conversational assistants from OpenAI, Google and a wave of well-funded startups that treat voice as the primary interface for human-to-AI rather than human-to-human conversation; these compress some of the same "talk instead of type" demand into a different use case. The third group is dedicated asynchronous voice products, a smaller field that has historically struggled to reach independent scale, with the most-cited cautionary case being Clubhouse's compression from a 2021 peak valuation back to a niche after pandemic tailwinds faded (public reporting, directional reference).

Where Haiket has a defensible edge today is narrow but real. Founder Alexander Narest's documented work on Google Duo's WaveNetEQ-based audio reconstruction and his prior shipping of Holler give him domain credibility on the hard parts of voice quality and on-device audio that a general-purpose consumer founder would lack [LinkedIn; AppAdvice]. That edge is perishable: it accelerates the first product version and helps recruit early engineering talent, but it does not create a structural moat against a WhatsApp feature release. Distribution, brand and graph effects all favor the incumbents.

Where Haiket is most exposed is the channel layer. WhatsApp does not need to win on context-aware voice; it only needs to ship a credible enough version inside the app users already have. If Meta releases an AI-mediated voice note feature in Western Europe in the next 12 to 18 months, the differentiation argument compresses sharply. The most plausible 18-month competitive scenario: Haiket wins if it converts the launch waitlist into a vocal early-adopter community in one or two European markets and uses that signal to attract a brand-name seed round, in which case it becomes an acquisition candidate for a larger communications or AI platform. It loses if waitlist-to-active conversion stays in the low single digits and an incumbent ships a comparable in-app feature, in which case the path to independent scale narrows.

Data Accuracy: ORANGE -- No competitors are named in the structured facts; competitive map is constructed from publicly known incumbents and adjacent products rather than from cited Haiket research.

Opportunity

PUBLIC

If voice genuinely becomes a primary interface for asynchronous personal communication over the next decade, the company that owns the default voice-first inbox in Europe could be a generationally important consumer business; Haiket is one of a small set of credible early entrants attempting that bet.

The headline opportunity. The single largest outcome Haiket could plausibly become is the default voice-first messenger in one or more Western European markets, a position analogous to what WhatsApp captured for text on mobile in the early 2010s. The cited evidence supporting this as reachable rather than aspirational is twofold: (1) John Donovan's launch-quote framing explicitly invokes the WhatsApp precedent and treats voice as "the next evolution of messaging" [PRNewswire, Nov 2025], and (2) the founder's shipped track record in voice quality (WaveNetEQ work in Google Duo) and consumer voice apps (Holler, with translation across 65 languages) suggests he has built the kind of product before [LinkedIn; AppAdvice]. The bet is timing-plus-execution rather than category creation.

Growth scenarios.

Scenario What happens Catalyst Why it's plausible
European default for voice-first messaging Haiket becomes the go-to standalone app for voice-native conversations in one or two EU markets, then expands Strong waitlist conversion in launch markets following the November 2025 release [Webull, Nov 2025] Founder has shipped consumer voice apps before [AppAdvice]
Acquisition by an incumbent platform A larger messaging, AI, or telecom player acquires Haiket for the team and the context-aware voice stack A demonstrated early-user metric attracts strategic interest Founder's Google and Fring network gives natural acquirer relationships [Yahoo Finance, 2025]
Embedded voice layer for third parties Haiket pivots or expands from a consumer app into a voice-experience SDK licensed to other apps A B2B partnership or developer launch The underlying audio and context-awareness IP is general-purpose [PRNewswire, Nov 2025]

What compounding looks like. A voice messenger compounds through the same mechanism a text messenger does: the address-book network effect. Each invited contact that joins makes the app more useful for the inviter, and each minute of voice content trains the context model on the platform's actual usage patterns rather than on synthetic data. Haiket has not yet disclosed metrics that would show this flywheel turning, so the compounding case is structural rather than evidenced today. A secondary loop, less dependent on social graph, is the data moat: a context-aware system improves with usage, so the first ten thousand engaged users disproportionately shape the model's behavior for the next million.

The size of the win. A credible public comparable for a successful European-rooted consumer messenger is Skype, which Microsoft acquired in 2011 for $8.5 billion (public M&A record, directional reference). Closer to the asynchronous-voice analogue, WhatsApp was acquired by Facebook in 2014 for approximately $19 billion (public M&A record, directional reference). If Haiket reaches even a small fraction of either trajectory, with single-digit-million European monthly active users on a voice-first product, it would be in the strategic-acquisition range for a major platform (scenario, not a forecast). The downside-bounded case is more modest: a sub-scale but profitable specialist serving a defined European user base, which still produces a credible founder outcome at this stage of capital intensity.

Data Accuracy: ORANGE -- Opportunity scenarios are constructed from cited launch materials and public M&A comparables; no Haiket traction or revenue figures have been disclosed.

Sources

PUBLIC

  1. [Haiket] Haiket | https://haiket.com/

  2. [Crunchbase] Haiket - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/haiket

  3. [Yahoo Finance, 2025] Technology Start-up Haiket Launches 'Context-Aware' Voice Messaging in Europe | https://finance.yahoo.com/news/technology-start-haiket-launches-context-134700530.html

  4. [PRNewswire, Nov 2025] Technology Start-up Haiket Launches 'Context-Aware' Voice Messaging in Europe | https://www.prnewswire.com/news-releases/technology-start-up---haiket---launches-context-aware-voice-messaging-in-europe-302611212.html

  5. [Investors Hangout, 2025] Haiket Revolutionizes Voice Messaging Experience in Europe | https://investorshangout.com/haiket-revolutionizes-voice-messaging-experience-in-europe-454259-/

  6. [Entertainment Newswire, 2025] Technology Start-up Haiket Launches Context-Aware Voice Messaging in Europe | https://entertainmentnewswire.com/press-release/technology-start-up-haiket-launches-context-aware-voice-messaging-in-europe-302611212-187613/

  7. [Webull, Nov 2025] Technology Start-up Haiket Launches 'Context-Aware' Voice Messaging in Europe | https://www.webull.com/news/13838705367966720

  8. [LinkedIn] Alexander Narest on LinkedIn: Improving Audio Quality in Duo with WaveNetEQ | https://dm.linkedin.com/posts/alexnerst_improving-audio-quality-in-duo-with-waveneteq-activity-7005805676237090816-npPh

  9. [AppAdvice] Holler - talk and text by Alexander Narest | https://appadvice.com/app/talknow-voice-calling/1575542140

  10. [Softonic] Haiket for AI Chat: review, features & use cases | https://haiket.en.softonic.com/android

Articles about Haiket

View on Startuply.vc