/Companies

Talkstack

Techstars '23

Singapore, SG · Founded 2023 · 2 employees on LinkedIn · 1 known investors

Find your way into Talkstack

Each arrow is one introduction. Sign up to fill in the first one.

14,079 people can intro you to Eunice

Ask Sagar Setu to introduce you to Eunice Wong

You →Sagar · together at Milky Way AI →Eunice

Sign up to find every way you can get introduced to Talkstack

  • Your email: who you already know at Talkstack, and who your contacts know
  • 2nd- and 3rd-degree connections: people you know who worked, studied or invested with the Talkstack team
  • Your LinkedIn: the connections who can make the intro for you

We rank every route by how warm it is, so you raise on warm intros instead of cold emails.

Talkstack AI provides an AI-powered agent platform for sales and customer support teams that operates via voice and text, automating complex workflows across multiple languages. The platform enables businesses to scale support and sales operations with human-AI collaboration and enterprise-grade security.

Also known as Talk Stack · TalkStack AI

Founders & leadership

Talkstack was founded in 2023 by Eunice Wong.

EWEunice Wong
Eunice WongCEO & Co-FounderShe has previously co-founded multiple AI startups backed by Y Combinator, TechStars, and Samsung.

Investors · 1

Techstarslisted by TechstarsBoulder · $220K

How we know: the Techstars portfolio · open dataset · Top 3000 Accelerator Startups (no YC) 2026-08 · Not right? Tell us

Company profile

researched Sep 2026

TalkStack AI (Talkstack, talkstack.ai) develops an AI agent platform that lets companies deploy conversational agents across voice calls, SMS, WhatsApp and other digital messaging channels. The agents are powered by large language models and are positioned for complex, multi-step task execution rather than rule-based scripting. Core use cases advertised include customer support case handling, lead qualification, appointment scheduling and autonomous reminders such as contract renewal notifications. A demo agent, "Joy", is exposed publicly via a US phone number on the company's homepage.

The product includes a no-code dashboard for configuring agents — cloning an existing agent's voice, style, tone and knowledge base — and a workflow builder that chains checkpoints across calls, wait times, SMS and WhatsApp reminders. Operational features cover call transcripts rendered as chat alongside audio recordings, response-accuracy review, analytics dashboards on multi-touch customer interactions, and graceful escalation to human agents either asynchronously or via live transfer. The company states support for more than 20 languages including major Asian languages, and cites enterprise-grade security and adherence to domain-specific SOPs with human oversight.

Published performance claims include handling up to 90% of Tier 1-2 support cases, up to 98% automation in live use cases, and a call completion rate above 98% for the voice agent. A third-party directory records approximately two employees on file.

Business model

B2B software platform sold to businesses deploying AI agents for customer-facing sales and support functions, with a free trial and a book-a-demo motion on the website; a third-party profile classifies the business model as B2B.

Traction

Customer testimonials cite inbound calling systems set up within days reaching thousands of customers, handling thousands of calls, and lead qualification at scale. Reported operating metrics include up to 98% automation in live deployments and over 98% call completion. Headcount on record is approximately 2.

▸Full profile — market position, technology, go-to-market, geography, risks & controversies

Market position

Positioned as an AI-native contact center stack intended to replace multiple point vendors, differentiating on agentic task execution versus rule-based competitor systems.

Emphasis on agentic multi-step task execution across chained channels rather than rigid rule-based flows, human-like localized voice with voice cloning, coverage of over 20 languages including Asian languages, no-code agent and workflow creation, and combined AI plus human agent handoff.

Technology

LLM-powered voice and text agents combining speech recognition and natural language processing to collect and categorize customer information and infer customer intent. The platform supports voice cloning, multi-agent chaining across WhatsApp, SMS and phone calls, a no-code custom workflow builder, call recording and transcript review, and analytics. The company describes its agents as agentic systems rather than rigid rule-based task chains, and cites enterprise-grade security controls. The public website is built on Webflow with GitBook documentation, Tailwind CSS, Cloudflare CDN, Google Workspace, Calendly and YouTube detected in its stack.

Go-to-market

Self-serve free trial and demo booking through the website, supported by customer testimonials from startup and enterprise users; the company is also listed in third-party AI tool directories and startup databases.

Businesses running sales and customer support operations, including both startups and enterprises; named or described users include NTUC First Campus, ShiftSmart, Yourlife.io and an unnamed top-10 insurance company. A third-party review notes the platform is oriented toward mid-sized and large businesses and regulated industries.

Geography

The website lists Singapore and San Francisco locations; third-party profiles place headquarters in the San Francisco Bay Area/San Francisco, United States. A quoted enterprise customer describes plans to scale the solution to 10 countries.

Risks & controversies

Third-party review commentary notes potential onboarding time for building customized workflows and voice cloning, a fit skewed toward mid-to-large organizations rather than small teams, and reliance on CRM or marketing automation integrations for maximum effect. Public data on the company is thin, with a low measured domain authority and a small recorded headcount.

Compiled by commissioned research from 5 cited public sources — announcements, filings, and press listed under research sources below.

Key figures

latest reported
Ahrefs Domain RatingSep 20268 score (0-100)
Automation rate in live use casesJan 202498%
Call completion rateJan 202498%
EmployeesSep 20262 people
HeadcountAug 20262
Languages supportedJan 202420 languages
Tier 1-2 support cases handled by AI agentsJan 202490%

Company-reported or press-reported figures, each dated to when it was claimed — not independently audited.

Related companies · 8

Hugging FaceHugging Face is a collaboration platform that hosts and provides access to machine learning models, datasets, and applications. It offers both open-source tools for the ML community and paid compute and enterprise solutions for teams building AI applications.
DataCampDataCamp is a learning platform for teams that teaches data and AI skills through hands-on coursework accessible via web browser and mobile app. It serves enterprise customers and development teams seeking to build technical capabilities.
DeepgramDeepgram builds foundational voice AI technology that enables speech-to-text, audio intelligence, text-to-speech, and voice agent APIs for businesses to create natural voice interactions. The company serves enterprises and developers across multiple industries seeking to integrate advanced voice capabilities into their applications and customer experiences.
Auth0Auth0 is an authentication and authorization platform that secures applications, AI agents, and internal tools by providing identity management, single sign-on, multi-factor authentication, and fine-grained access control capabilities.
ResendResend provides an email sending platform for developers to build and send modern email templates that reach the inbox reliably. The platform addresses the technical challenges of email delivery across different email clients.
Zendesk New YorkThis is Zendesk's internal belonging/diversity and inclusion program page, not a company profile page for an external company. It describes Zendesk's internal culture initiatives and leadership development programs.
NebiusNebius operates an AI-focused cloud platform offering GPU infrastructure, MLOps tooling, and managed and serverless inference for AI training and deployment. It serves AI developers and enterprises, providing custom hardware with non-virtualized NVIDIA GPUs and InfiniBand for workloads ranging from experiments to global-scale environments.
ZapierZapier provides a platform for connecting AI models and agents to workflows and business applications, enabling teams to automate tasks across over 9,000 apps without coding. The platform serves enterprises, operations teams, and businesses of all sizes looking to deploy AI-driven automation for lead capture, customer support, employee onboarding, and other business processes.

Companies working in the same space as Talkstack.

▸Research sources · 5

primary sources listed

5 public sources were cited for this profile; the first-party ones are listed here.

Frequently asked questions

What does Talkstack do?
TalkStack AI builds LLM-powered voice and text AI agents that automate sales and customer support workflows.
Who founded Talkstack?
Talkstack was founded by Eunice Wong in 2023.
Who are Talkstack's investors?
Talkstack's investors include Techstars.
Where is Talkstack headquartered?
Talkstack is headquartered in Singapore, SG.