Agora's Voice AI Agents Hit Their Stride: From POC to Production
Second-quarter revenue accelerates 18% as conversational AI moves from pilots to commercial deployments; net revenue retention rebounds past 100%.
API · Earnings Call · 2026-08-13
A Pivotal Quarter for Conversational AI
Agora's second-quarter results mark a clear inflection point. Revenue reached $40.4 million, up 18% year-over-year—the third consecutive quarter of accelerating growth (component 3356771550991698041). The engine is Voice AI agents, which are finally moving from proof-of-concept to commercial production. As CEO Tony Zhao put it in prepared remarks, “customers are now seeing substantial cost savings from deploying our voice AI agent.” — Bin Zhao, Executive (likely CEO or similar) based on leading remarks and detailed business discussion · 2026-08-13 This is no longer a speculative narrative; the company delivered its seventh consecutive GAAP-profitable quarter, with net income of $2.2 million, up 50% year-over-year. The core Real Time Engagement business remains stable, but the growth is being driven by the AI products. Perhaps the most telling metric is the dollar-based net retention rate, which jumped to 104% from 94% a year earlier. CFO Jingbo Wang highlighted the recovery, noting that it "represents a meaningful improvement and moves us back above 100%" (component 3356771550991698041). This is a strong signal that existing customers are deploying AI features at scale, not just testing them.From POC to Production: The 5% Milestone
The target of 5% revenue contribution from conversational AI by year-end is not new—management has been guiding to it since early 2026. In the May 2026 call, Jingbo Wang said, “we expect to see revenue to quickly ramp up and possibly in towards somewhere around 5% revenue contribution by year-end.” — Jingbo Wang, CFO · 2026-05-26 What has changed is the conviction. On this call, the same target is stated with more confidence, and the use cases have broadened: outbound marketing, market surveys, financial services outreach, and debt collection. Tony Zhao described how the voice AI agent now handles outbound marketing with conversion rates comparable to human reps, and can compress weeks of survey work into hours. The company is targeting $41-42 million in Q3 revenue, implying 15.8-18.6% year-over-year growth—and still ramping.Competitive Landscape and Infrastructure Advantage
Competition is heating up, with Twilio and others attacking the conversational AI space from different angles. Tony Zhao acknowledged the race but positioned Agora's edge: “Given the huge potential of the conversational AI market, it is natural to have competition.” — Bin Zhao, Executive (likely CEO or similar) based on leading remarks and detailed business discussion · 2026-08-13 He emphasized that much of the technology involves audio processing—handling noise, echo, and packet loss—where Agora has deep experience. This infrastructure moat is complemented by new developer tools: Agora Skills and Agora CLI, which make it easier for AI coding agents to build on the platform. The partnership with Gradium to integrate TTS without added latency further strengthens the ecosystem.What's New, What's Contested
The focus on call centers is a continuation of a theme Agora has been building since early 2025. In the March 2026 call, Tony Zhao described a similar scope: “We're now also focusing closely with global customers from U.S., Europe, South America, Asia Pacific region and inside China to implement our solution in a couple of customer service scenarios.” — Bin Zhao, Chief Executive Officer · 2026-03-03 What is new this quarter is the clear evidence of production adoption—retention above 100%, consistent GAAP profitability, and a CEO personally committing to buy $20 million of additional shares once blackout restrictions lift. The company has already returned $160 million to shareholders through buybacks, a notable point given its market cap of roughly $320 million.Agora is betting that the scarcity in the AI stack is not in the models themselves, but in the real-time, low-latency infrastructure that makes voice AI agents actually work in production. With a strengthening balance sheet, accelerating revenue, and a clear path to scale in conversational AI, this quarter is a definitive validation of that bet.We believe the center of gravity in the AI industry is increasingly shifting from model capabilities towards infrastructure and hardness layer required to operate voice AI agents reliably at scale.